Oracle first came into Japan more than 25 years ago, but the challenges they faced and overcame then are exactly the same ones firms are facing today in executing their Japan market entry.
Allen explains why Oracle needed a unique sales and marketing strategy for Japan, and how he managed to get buy-in from headquarters — even though Oracle already had a sales and marketing program that had proven fantastically successful in other markets.
We also talk about how Oracle managed to negotiate a amicable exit out from their exclusive distribution agreements not just once, but twice. That’s an amazing accomplishment considering that many foreign companies have destroyed their Japanese business the first time they attempt it.
But Allen, tells the story much better than I do. I think you’ll enjoy the interview. I know I did.
Welcome to disrupting Japan. Straight talk from Japan’s most successful entrepreneurs.
I’m Tim Romero, and thanks for joining me.
Japan is slowly opening up again. The official “unofficial” lockdown ended at the beginning of June. Restaurants, bars, and shops are reopening with a lot if plastic curtains and sheeting separating patrons and proprietors. It’s a long way from normal, but it’s better than being stuck in the house.
International travel is mostly shut down, but domestic travel is really picking up. It seems most of the hotels and resorts in Okinawa are already booked solid for the summer by Japanese who would normally be flying to Hawaii. And Okinawans, grateful for the business, but still nervous about the virus, have some pretty mixed feeling about that.
And of course, with international travel shut down, and all the trade shows canceled, most foreign startups have put their Japan market entry plans on hold. And that’s normally a lot of activity. If you are a B2B startup you need to be looking at Japan. It can be a hard market to crack, but it’’s a lucrative one.
So today, I want to re-share what is one of the most amazing Japan market-entry stories of all time. It has ambition, misdirection. drama, serious career-risk, and rock-concerts. It’s an old story, but a good one. The technologies have changed since then, but the challenges and the strategies haven’t.
To kick things off today, we’ll get a chance to sit down and talk with my good friend Allen Miner about the challenges Oracle faced, and overcame, when breaking into Japan.
I’ll warn you in advance that this episode is longer than most, and believe me, I cut things to the bone. But there is just too much great information about how to overcome both the personal and professional challenges that foreign companies face here. I felt like I would be cheating you if I edited out any more. In fact, Allen explains how Oracle successfully maneuvered out of an exclusive distribution agreement, not only once, but two separate times. This is something that has sunk more than one foreign company here. But Allen tells the story much better than I can, so let’s get right to the interview.
Tim: So I’m sitting down here with Allen Miner and Allen, you’ve been involved with the market entry of a lot of companies into Japan. But today I want to focus on the one that you led personally, which was Oracle Japan. So let’s back up. What was attractive about the Japanese market? What made Oracle decide that they needed to be in this country?
Allen: Actually, that happened a few years before I joined Oracle. In, I believe it was 1982, Oracle was about a $5 million a year company worldwide, 5 years old as a company, and just released their first commercial version of the Oracle database software. There was quite a bit of press about, “How interesting is this relation to technology? It doesn’t require traditional programming to do data manipulation.” And the U.S. press got read by some technical geeks in Europe. And one in particular in Japan said, “This sounds really interesting. We ought to figure out if we can bring this cool new technology to Japan.”
Tim: So it was a partner company pulling you in?
Allen: Yeah, it was a company called Digital Computers Limited, that at the time was building DEC VAX clones. Because Oracle originally was released on the DEC VAX computer platform, the president of that company, a gentleman named Mr. Yamada had read an article about it. So this is when we reached out to Oracle and see if we can sell their software in Japan and that contact from an interested Japanese distributer was what got it all started back in 1982.
Tim: Okay. Even today, I think that still a really common case.
Allen: Yeah. I think it is. For companies that come into the market early, perhaps earlier than they are really ready, I think that’s the most common. It’s very common for a young company with really interesting technology to be found by someone in Japan, or other countries of the world. Everyone knows that Japan is potentially a big market. I don’t think there’s a question of the potential opportunity or the size of the potential market in Japan, but often it is the trigger of, “Oh, someone is interested in our software in Japan.”
Tim: But it seems like that could really be a two-edged sword. Oracle was not a small company at that point.
Allen: No we weren’t. It was 5 million in revenue, maybe 60 people. There was no international division at the time so the person who took the inquiry was one of, I think, 3 or 4 U.S. sales representatives.
Tim: How did you guys vet this company? How did they make sure it was for real?
Allen: The president and one of his technical staff, I understand, flew to Oracle’s headquarters in California, met with the sales team at length. I’m sure that because they were clearly well-informed about the DEC VAX environment, they had customers for their computer products, they clearly had some kind of understanding with what you could do with a relational database that I think was some technical vetting, maybe not a lot of time spent on what they might be able to do in the marketplace. I’m not sure that Oracle really, at that point, had really figured out how it was going to grow the market.
Tim: So it just was a good strategic fit and it was a market they couldn’t have addressed anyway, so let’s give it a try.
Allen: They weren’t even actively pursuing international opportunities yet so I think our initial entry into England, the Netherlands, and Japan all started that way, with some local geek who was always staying on top of the latest technology trends reaching out and saying, “We’d love to distribute your software in our country.”
Tim: All right. So things obviously went well for them. They sold the product.
Tim: How did you get involved? When did they bring you on board?
Allen: Well, as Oracle continued to grow, we expanded the platforms that the product ran on. It was initially on the VAX, then we introduced UNIX platforms, the PC, and even IBM mainframe computers. As we were expanding the platforms for Digital Computer Limited, the UNIX environment and the VMS environment were quite comfortable. But when we wanted to introduce a mainframe product, they didn’t know anything about the IBM space, they didn’t want to get involved with the IBM space, and by that time we had hired a vice president of international. I think he had one or two staff and their company was maybe a couple hundred people by then, worldwide. The decision was that we were now proactively trying to grow the business internationally and if the distributor in a particular country was not interested in the mainframe product, we believed at the time we were going to have a huge business in the IBM mainframe world. Turned out not to be the case but at the time that was what the folks in Oracle believed and because Digital Computer Limited didn’t understand, didn’t want to pursue the mainframe space, the decision was made to identify a second distributor in Japan that would focus on the mainframe products. This was in 1986 that this work was going on—or 1985 rather—and we added a second distributor called NESHEEN Products that had experience with the product out of Germany.
Tim: So for the structure of this, did each of these distributors have an exclusive right to resell the product—
Allen: So Digital Computer Limited had exclusive rights around the VAX line of computers. NESHEEN Products had exclusive rights around the mainframe and both companies were free to sell the UNIX and PC products.
Tim: Okay. This is a challenge I think a lot of companies get themselves into coming into Japan, is if you give a distributer exclusive rights, and they’re successful, you’re going to want to come into the market yourself and that’s exactly what you did.
Allen: That’s exactly what happened with Informatic.
Tim: How did you deal with that? How did you kind of change the game for these exclusive arrangements you had?
Allen: Well, I think—remember we had the two distributers in Japan with distinct sectors of the market. I think between the two of them, the fact that we had those two distributers and we did not have a Japanese language product at the time, and the two distributers apparently were arguing about what the specs should look like and which of them should be authorized to build it for Oracle. Which of the two were technically more competent to advise us on building a Japanese version. That is the situation in which Oracle came and interviewed me on the campus university, and when they noticed that I was not only a computer science student, but I spoke Japanese, the recruiter said, “We want to hire you.” He said, “Oh, I see you speak Japanese.” I said, “Yes, but why is that relevant in a computer programming giant?” And he explained the situation, “We need someone to sort this out for us.” And it was the first time I realized that there was such a thing as making software work in the Japanese language.
Tim: Right. It was a lot harder back then.
Allen: And no one had really figured it out yet. And then the caliber of the people attracted me to Oracle and my first job was basically to meet with the distributers from Japan and form my own opinions about their technical capabilities and develop and resolve together with them what the Japanese specs should look like, and then build and develop the Japanese product.
Tim: So did you end up going with one or the other to implement it or did Oracle do it in house?
Allen: What I concluded was that the Digital Computer team had a clear understanding of what should be done and how to do it, and were further along in actually building their own prototype of a Japanese version. So I fairly quickly made the choice to collaborate with them on developing the specification but Oracle was very uncomfortable with the idea of outsourcing that development to a third-party. So my job was basically to work with them, have them provide me code samples they built, and then within—although I was working in the international division, I also had sort of a dotted line database development team that gave me access to source code. And I was able to sort of build the prototypes and check in Japanese code into the source control system.
Tim: Okay. That’s almost an ideal working relationship with your distributers. Was Oracle thinking about setting up a direct sales force here in Japan or were they planning on just continuing to use distributers?
Allen: This was 1986, in a time we weren’t. I think in that year that I joined, the total revenues from Japan were $250,000 out of about $25 million of revenue worldwide for Oracle. So still small, but NESHEEN Products had just got going around with the mainframe product. Toshiba had also become an OEM by that point and was reselling the product on their UNIX computers, and Digital Computer was selling some. My task, as well as building the Japanese language product, was to figure out how, with the distributers, to grow the revenues to half-a-million dollars in that first year. Fresh out of college, working at headquarters, and we got to about $495,000 in revenues with a new OEM relationship and increased sales with the partners. There was a very specific event that occurred that accelerated our transition from . So, when I joined I think the intention was to continue working with those distributers and just continue growing the business organically but one year after I joined, the international VP received a call from Bill , who runs the Assisto software distribution company in Japan. And the two of them had worked together a few years before our VP joined Oracle and Bill decided he wanted to get into the relational database space. He wanted to have the relational database part excel and he discovered that Ingres did not have a distributer in Japan yet. Looking back, most people remember the battles between Oracle and Sybase for dominance of the database market. Actually, in the early days, the battle to the death was between Ingres and Oracle. So in 1987, we were still very much in the middle of trying to break away from Ingres. Both companies had just gone public within the last year. Both were working toward $50 million revenue years and Bill called my boss and said, “We’re thinking of distributing Ingres in Japan. What do you think? What do you think of the product?” And kind of getting friendly advice, and obviously getting it from a competitor, he knew he was going to get any of the warts that he’s going to get told about. Immediately after that call, a little bit of a war room meeting was pulled together where John said, “ASHISTO has such a powerful sales organization and they are in all of the major accounts in Japan, through all their older mainframe software products, that if they take Ingres while the products are still competitive and are competing against our two distributers in Japan right now, we will lose the market.” So we went into immediate war room; his answer was, “Ingres is a good product but so is Oracle. We’re not sure that we’re entirely happy with our distributers. We’d like you to come and take a look at Oracle as well and compare the two before you decide which one to sell.
Tim: Now, from headquarters’ point of view, that’s a great decision. It’s kind of an easy decision, especially if you can convince them to sell Oracle instead, but from where you were sitting, you’ve got a bit of a challenge ahead of you if you’ve got two existing distributers who have exclusive rights to particular platforms—
Allen: And I spent a year building a personal relationship with them, around building a trust to—
Tim: How do you dance around this? How do you do this?
Allen: Well, the first type of course was if we were not able to win over Assisto to our side, it was not even a question. So the first step—we were convinced it would be disastrous if Assisto carried Ingres and not Oracle, and that it would be very good for us if we could get them to distribute Oracle. So the first thing we did is, fortunately, I had a year working on the Japanese language stuff, spoke Japanese. Oracle was slightly ahead in the market at the time and had a well-deserved reputation for being, between the two, the stronger sales and marketing organization. And Bill sent a couple of engineers over for a week, spent 2 or 3 days over at Oracle, 2 or 3 days over at Ingres, and went back with the recommendation to Bill that the Ingres product looked like it was technically better. But Oracle’s sales and marketing organization seems much stronger and they have an engineer who can speak and support us in Japanese, and they would be working on Japanese language support. So the feedback we got from Bill, when they decided which one they would prefer, if possible, to do a deal with us, is that he believes in the end, the stronger sales and marketing organization will win in the technology space. The technology matters but sales and marketing matters even more. He was going to put his bet on the stronger sales organization, not the stronger technical product.
Tim: And that’s pretty much how it played out too. So he decided that to carry Oracle exclusively and not carry Ingres?
Allen: He did. And so we then began privately negotiating with Assisto to figure out if we explored potentially organizing it as a joint venture between the two companies—explored that for a bit. Negotiated the basic terms of the contract. Bill told us what kind of terms he needed in order to run a profitable business, and invest for its growth, and he wanted a lower royalty rate than we were accustomed to giving, so John came up with the creative solution that we would create an international price list that sets the international price at the level where your royalty rate works for you. But we’ll price it to you on the standard percentage of the U.S. price list.
Tim: So about how much more—what percentage more was he—
Allen: The Oracle model at the time was generally to give 50% to the distributer and keep 50% for itself. And the contracts were all based off the U.S.’s price, which Oracle controlled and could change whenever they wanted to. So the royalty rate to Oracle was always 50%. And what John did was help create an international price list that was 150% of the U.S. price list and we’ll set that benchmark so that when you have customers asking what the price is, you can give them the Oracle international price list, and the royalty that you pay to Oracle, even though it’s in the contract 50% of the U.S. price, you can charge them whatever price in the market, we’ll give you some materials to help testify our price, and you can build a model where you get the 33%.
Tim: Do you think that strategy—that strategy was much easier to execute back then but do you think that it’s still viable in this era of sass, and online marketing, and pricing information? Do you think companies could still do that?
Allen: The short answer is yes. I’m not sure that you can carry the same discrepancy in the price that you used to be able to but buyers in a country will always prefer to do transactions in local currency. And at least on the level of, we need to have a higher price to allow us to absorb currency fluctuations and there’s some willingness to accept that as a premise for a higher price. Also, if you want to buy the U.S. product from the United States, and have no access to support in Japan, and always do the , yes, you can do that.
Tim: That’s true. Right. Local support is really critical.
Allen: Local support, the Japanese language capability was not yet a standard feature of the product. If you want a Japanese language product, you have to buy it from Assisto, so there are a number of things that gave them pricing—but I think you’re correct. It’s more difficult. I’m not sure that I’m aware of any companies recently that are able to maintain a 50% premium over their U.S. price in foreign markets.
Tim: Okay. So Assisto was successful. They were aggressive in selling the product into the market. What was the reaction of your existing distributers?
Allen: Well once we’d actually sort everything out and have the contract done with Assisto, we knew that we had an out. One was that the mainframe distributer had quotas within their contract that they had not met so we had the right to cancel the contract for failure to meet the agreed upon sales targets. And the Digital Computer Limited were about a half a year away from a five-year optional renewal, so we just went in with them and told them that we were not going to renew when the five years came up and if they were willing, we would like to simplify things in Japan by ending the contract immediately. This was about a month after we made our arrangement with Assisto and while they were still preparing their team and getting trained, we were going to go to the market. The transition, timing-wise, would have worked out just right.
Tim: So did that end the relationship or did they become resellers under Assisto?
Allen: Well, that’s interesting because in the meetings that I had with my boss from Oracle, sitting down with him, the tone was very aggressive, no room for negotiations, no room for discussion, you missed your quote, I’m not going to listen to any excuses, lack of Japanese or whatever it is, you missed the quota, letter of the law, we’re out. And the tone of the meeting, for a young and friendly engineering type like me, it was very uncomfortable to be in a meeting with my friends hearing my boss take this very hard line with them. In that same week, though, I forget if we had planned it in advance or if in conversations—a gentleman named , who was a director at Assisto, and was the executive in charge of the Oracle relationship, he and I went very shortly thereafter, maybe even immediately thereafter, visited both of those distributers again, and he said, “I understand and appreciate the investment you’ve made in understanding Oracle, and building a pipeline in having customers, and if you will accept it, I would like you to continue to be a second tier distributer under us, servicing your customers, building the business, recouping whatever you’re invested in learning. And he went in and basically made a very sincere effort to smooth things out.
Tim: Did they end up staying on?
Allen: Both of them did for a while, but NESHEEN Products have not really built enough momentum in the business for it to be that critical to them. Once they had finished closing the business they were already working on, they wound down their Oracle business, but Digital Computer Limited, for the next 20 years, continued to be a very committed, loyal partner of Oracle. I don’t think they ever got more than about a billion dollars a year in sales, so the decision to switch when they were at about a quarter of a million—it was a good time to switch them, but they remain very loyal, very technical, good partners for Assisto and Oracle for many, many years.
Tim: I think this illustrates one of the biggest differences between U.S. sales culture and Japanese sales culture. U.S. sales culture tends to be very transactional in nature. If you miss your quota, you miss it. You’re done. Like you said, it’s the letter of the law where Japan is very relationship-based. Was this planned going in or was it just the U.S. side—that’s the way they are and they don’t really care about—if things are done differently in Japan, work it out in Japan, but—
Allen: First I’m going to react to that and say that I think that characterization is true as a stereotype. That criticism is stereotypically true, but subsequent Oracle, we’ve brought additional companies into Japan, Salesforce, Concur; one of the key things that I have looked for in choosing a company to bring into Japan, of course, the momentum they have in the United States, the quality of the team, the likely success with a U.S. company is critical. But I have met with numerous companies that I have chosen not to pursue a relationship with, specifically because my perception is that the executives have that, what you characterize as a very American transactional attitude toward their business. And I have always opted to engage in a company where the executive leadership seemed to have more of a relational, long-term relationship, mutual benefit, a balanced approach to building powerful, long-term relationships in their attitude and approach to engaging with us, partly because I think it will be better for the Japanese partner, but mostly because those are the people I would rather work with. And I’ve also seen Japanese companies and Japanese people who behave in a very transactional American way. I think that’s stereotypical, but in my experience, I would say, in general, the companies that are successful in Japan are ones that either has a culture have more of a long-term building value together relationship approach to how they engage with partners and customers, than a very straightforward money-for-value transactional approach. So ones where that’s kind of built into the culture, where alternatively, Oracle certainly did not have that kind of culture but my personality was such. And so in the case of Oracle, I think the fact that my orientation was that way and I was the person assigned to think about and work with folks in Japan, was helpful when there were transactional kinds of conversations, for me to go back and pre- and post- those, manage the relationship and show the long-term commitment of me, as a person of Oracle, to the relationship, even if sometimes the conversations with people from headquarters were more transactional.
Tim: That makes sense. So if there was a necessary shift in style between headquarters and Japan, were there any changes to the marketing or the product itself, beyond just localization?
Allen: That’s actually an interesting one. One of my biggest frustrations in the early days of Assisto is that Oracle had developed in the United States a very finely-tuned go-to-market strategy, which in the software industry today is almost common sense. At the time it was very innovative. Tom Siebel, who later started Siebel Systems and created the software category of CRM, at the time in Oracle, was in charge of the telesales and telemarketing group. And Oracle has a program where we would send out—it was pre e-mail days so we would take advertisements in the industry press talking about 30 of 40 seminars we were going to do around the United States and to take registrations by telephone for these seminars. We would send out invitation letters to people on the mailing list of these magazines, so we had a program to cause people to call in and sign up to an Oracle introductory seminar. And when they would call in, we would ask fairly extensive qualifications questions to the point where if I had been on the receiving end of them, I probably would have ended up hanging up halfway through Assisto. It’s like, “Do you have budgets allocated? What other products are you looking at? Who’s your decision maker? What application are you looking at?” This was all in order to be able to register to come to a seminar where they’re going to tell you about the product. So there was a very thorough pre-qualification process before anyone ever sat in a room and heard the general Oracle story, so that when people came to the seminar, we knew who in the room was a serious prospect who was also looking at a competitor. We knew who was there just to learn and didn’t really matter from a sales point of view. You would have very focused conversations in the break times with the people that you knew with a highly probable—this whole process of how we used advertising, direct mail, telemarketing qualifications, seminars, telesales follow-up, field lead allocations, the whole kind of lead pipeline management process that’s very common in the industry today. As far as I know, it pretty much originated with Larry and Tom Siebel in the 1980’s. Joining Oracle fresh out of college and having heard over and over again about this wonderful program, the Oracle way of marketing, I was totally bought into the Kool-Aid. And I still think it was a very creative and very efficient way to build a sales pipeline. I think that process itself was almost certainly fundamental to Oracle’s ability to double its revenue every year for 11 years. The efficiency of that whole pipeline sales process was a key element of that growth, but Assisto had a different way. I had lived in the United States and had come to Japan on business trips for the first year when we were working with two distributers. When we decided to have Assisto become our distributer, I had seen enough and heard enough excuses from the previous distributers about why they weren’t selling more. And in each of the meetings, when we were cancelling the contracts, a number of reasons were given, “This is why we’re not selling more,” to try and persuade John that we should continue with the distributers. What I realized in those meetings is that neither John nor I had a clue which of the reasons they were giving us for Oracle’s failure in Japan were serious problems, that if we did not resolve them, would make it impossible for us to be successful even with our distributer Assisto, and which of them were just excuses for their own inadequacies. So I had no basis and John had no basis upon which to evaluate the issues that they were raising, about why we’re not selling more in Japan. So in the context of negotiating with Assisto, I said I know when I joined the company I said I don’t want to live in Japan, I just want to with Japan from California, I think I should move to Japan for a year or two at Assisto and understand what’s really going on and treat them almost as if they were a subsidiary, give them access to information, and kind of be part of the team while they’re getting off the ground, so that—
Tim: I think that is a problem so many companies still face with their Japanese subsidiary. It’s a bit of a black box. Until you move here and actually are selling it yourself, it’s very difficult to know what is a legitimate cultural or market difference and what is just salesmen making an excuse.
Allen: And sometimes there are things that are legitimate problems that you can push back and resolve them locally. Sometimes there are things that need engaging headquarters in order to be resolved so I think that’s true. It’s very difficult, I think, for any branch, even if it was Sidney or London. I think the physical remoteness and the infrequency of direct contact, even in a similar business culture and same language culture, it’s often difficult to understand what the real issues are far away from that office. But with the addition to differences in business culture and language and—it makes it that much harder with Japan to tease out those differences. You either have to hope that you’ve hired a competent leader, or engaged with a partner, and take what they say at face value, and trust them and do what they say they need, or second-guess everything they tell you and neither is really the truth.
Tim: So when you came here, did you try the Oracle playbook in Japan?
Allen: Back to that question, so once we gotten Assisto on board, I had heard what a great sales company they are, they clearly seemed to understand how you go about selling software. I moved to Japan and physically located myself in the office with them and what I wanted to do more than anything was trying to get them to do this Oracle way—this tried and true Oracle way of marketing. And I spent a fair amount of time explaining how we do it, why it works, why it’s so wonderful. And Japanese magazines don’t give out their mailing lists, so we can’t sell direct to people in their mailing list so that was the first thing. So I told headquarters, I said, “Well, we can get it from Computer World. We spend so much advertisement on Computer World, they’ll meet us anywhere in the world and I’ll bet you can get it.” So I called and sure enough, we could get the Computer World mailing list. He says, “Well, Computer World will do it for full price,” and that objection was ticked off. So I think we did some advertising in Computer World and did a mailer to them. And another part of the Oracle way was that it was done in major cities all around the world, so people could go to a local hotel ballroom and hear about Oracle. I said, “We should go to . We should go on this road show and every quarter, in each city. We have this lovely office here, right here in , and there’s so many people here in Tokyo and we would like them to come to our office and see that we’re a credible company, and get comfortable with buying from us.” So we wanted them to come to us and the Oracle way is you want them in a neutral place where they don’t feel the pressure of being in your office in that initial encounter. So we ended up doing it in the Assisto adapted way but for the 3 years that I was working with them, we tried—we did get them with pressure for a year or so, I think Japan was reporting into Australia and Australia, better than any other country in the world in Oracle, really, really lived by that. And the managing director from Australia said, “I’m going to come up to Japan, we’re going to do it the right way,” so we rented a hall in a Hotel Okura, did it one time the way he said, but then he was off and—their method was cold calling, lots of face time, visiting customers, visiting their existing customers—
Tim: So now this is interesting. They were—
Allen: Very interactive, face to face, visit your office—
Tim: But they were getting results with this, right?
Allen: They were. So one of the other things that was really puzzling to me in the early days with Oracle is I was trained in my new employee training how to ask a customer what they’re trying to build with the product and develop overnight a partial prototype of what they’re trying to do, so that as assistant engineer, on your second call with a customer, you actually show them, with some sample data, and a few data entry screens and reports, what they could do with Oracle addressing the problem they specifically want to address.
Tim: So they think, “Wow, he did this overnight? This is easy!”
Allen: Yeah, it’s going to be very productive and that was the way I was trained to sell. I tried a few times to go along with the sales reps from Assisto. I would go on calls together with them and I would go in and the sales rep would say, “Hello Mr. So-and-So, I’m with Assisto, nice to see you again. We’re carrying this new product called Oracle now. It’s relational database. I’m not really sure what it does but would you please buy one?”
Tim: And it worked?
Allen: Well, obviously overstating or simplifying it but essentially, he would go into people that he’d been visiting for many times, had sold software to before, and because Assisto had such a good reputation for support and his products, he would just go in and beg for an order, and a few weeks later, he would have an order. I had not done any demonstrations, we had not had any conversations about what they should do with Oracle—
Tim: So it was purely the strength of the relationship?
Allen: The strength of the relationship and the trust that they had for Assisto was so powerful that they had a salesperson who had only been selling to IBM customers. His first problem was, he had no idea where are the VAX’s in Japan, so this is the first time Assisto was doing something out of the idea mainframe space. And because DEC computers were mostly used in factories and laboratories at the time, Assisto had no connection. So part of his challenge was figuring out how to get into the departments he had never sold to, with customers where there was somebody who he could verify that Assisto is a credible, good company to buy from. So that was stunning to me that going in and begging for the business—
Tim: But looking back in retrospect now, knowing everything you know now, Assisto’s Japanese methods are yielding results. They’re getting sales, it’s growing, headquarters has their way. How much of this conflict was simply each side wanting to do things their way and how much of it was really, “You should do what works and this will improve things?”
Allen: I think there was some of both. And I think in the end, the reason that it was frustrating—but it never got out of hand that they weren’t doing the Oracle way. It was never close to being a relationship-breaker for us. It was frustrating because we all believed they would be so much more productive and so much more efficient if they would do it that way. But they didn’t want to and in the end our basic operating principle at Oracle at the time was, “How much did you sell last year? Multiply it by 2. That’s your quota for next year and figure out how to do it.” And because Assisto was successfully doubling in size every year and meeting the commitments that they had made in the contracts to us, doing it their way, and because Oracle was growing well enough and Assisto was building a good enough relationship for us in the marketplace, we were able to continue adding relations with Fujitsu, and NEC, and Hitachi, and continuing to build out the hardware manufacture system integrator relationships that we wanted to, we were able to keep growing the business by 100% every year. So the results, by delivering on their commitments, Assisto gained quite a bit of trust and flexibility, even though I’m not sure I ever understood why it worked as well as it did for them, and even though right up to the day that we moved to the subsidiary stage of our business, I remained frustrated that I couldn’t find the secret sauce to convince them that the Oracle was would be so much more productive for them. The reality was—
Tim: Did you find that you spent more of your time explaining to headquarters why the Assisto way was working and you should let them kind of customize it or did you spend more time trying to convince Assisto that the Oracle way is the way to go and you should really give it a chance?
Allen: No. I think I spent most of my time making sure that Assisto felt like they were part of the family. That whenever I would go to headquarters, I would wander around and pick up every piece of new collateral or marketing information that was available. I took a very open attitude to how much information I would share with them, at what stage, and from my point of view, the more I could treat them the way we treated our subsidiaries, the more likely they would be able to be successful. Most of the time I spent was making sure that they got access to what they felt like they needed, that I was proactively telling them what was coming down the line so they could be prepared.
Tim: Relationship building.
Allen: Yeah, it was basically relationship building. And then because I was this Japanese-speaking American in a day when there weren’t that many in Japan, they’d have me go out and do seminars, and things with customers as sort of a—Bill’s not the only strange guy we have. We’ve got another one from Oracle too.
Tim: Okay but now with this success, Oracle is about to make another transition. You’ve transitioned away from your original distributers and that worked surprisingly well and now how do you transition from the distributor into a subsidiary arrangement?
Allen: Well, let’s start with “Why?” We had gone from $250,000 the year that I joined Oracle to about $10 million a year four years later. Assisto had come to the end of their initial contractual commitment of, “We’re going to do a million in sales, then 2 million in sales, then 4 million,” and we were trying to get their heads around how we could double the size of their organization from 60 people to 120 in one year, and how they could go from 4 million to 8 million in revenues in one year, and we felt that part of the way they would do that is begin to carry the Oracle accounting package and the Oracle manufacturing package. So Oracle, by then, had just released the first version of its applications packages. As we were going through the discussions with Assisto about, “What’s next after what’s in the contract?” And they had had a very unpleasant and unsuccessful experience with an accounting package software for mainframes previously and were convinced, as I think many people still are in Japan, that Japan is predominantly not a package software world, but a custom built software world, and that even many of the SAP applications are highly customized in Japan. Assisto was convinced with significant evidence, that I think is still significantly true today, that generic application packages were too hard a business to get off the ground in Japan and so as DCL had originally not been interested in the mainframe products, Assisto was not interested in carrying the application products.
Tim: Was that the leverage or the excuse you kind of used to—
Allen: So I had, from the day I began the relationship with Assisto, I had always been telling them, at some point in the future in Oracle’s natural evolution, we will create a subsidiary in Japan. And our international strategy will always have that as an endgame.
Tim: So you were managing those expectations from the very beginning?
Allen: From the very beginning, I said, “This is just the way we work and I want you to know is that what I hope happens is when that time comes, we’ll all know each other so well that a very real possibility is that we work with Assisto to have this organization become the Oracle subsidiary in Japan, if that’s mutually interesting. Alternatively, what I promise to you is that I’ll see you get treated the same way that you treated DCL and NEESHEEN when we switched to you, so that you can continue to build your business in the Oracle family after we do it.”
Tim: And were they amenable to that?
Allen: I think the honest feeling is, “Yeah, go ahead and do it when you want. tried to do it and failed. We’ve had many companies try to leave us and fail. You’ll just be less profitable and you won’t be able to do it as well as we can.” So I think there had been enough cases where companies had done that and it had not worked out well, and they had regretted it—
Tim: Well, it makes sense in Japan. They own those relationships, right?
Allen: They do. And very, very often the transition from distributor to subsidiary is handled badly and kills the business. So I think on one level, they wondered how serious we were; on another, I had been assuring them that I wanted to continue the relationship that I had spent 3 years building—the personal relationship with all of them, but what we also did, I had informed them when I began interviewing candidates for the new subsidiary CEO role—it was about a 9 or 10 month process before we finally found someone that was the right guy—but the way that I had envisioned coming into Japan without introducing too much conflict in the marketplace initially was that Oracle believed the applications business was going to be as its database business. I sure didn’t want to do that. They had a good growing database business. This transition was all about, “What can we do to get the total business in Japan to double? And even if the database business grows only 60%, if we can add applications on top of that, and as long as the total business is doubling, I’m doing my job.” So the basic premise for creating the subsidiary and hiring the CEO was to create a subsidiary that would roll out the Oracle applications in Japan and not be able to conflict with Assisto.
Tim: So they didn’t have any stake in the subsidiary at all?
Allen: No. We talked about it a little bit but they did not. Because in our early conversations, we had talked about maybe setting up the original distributorship with the joint venture, we had allowed them to use the name for their wholly owned subsidiary, that was a distributor for Oracle. So there was an interesting twist; we had allowed them the leverage the brand as if it were their own in building the business initially.
Tim: Now, did they transfer that back to you out of good will?
Allen: Oh yeah. Ultimately, we had did. Yeah. One year after the search started, from IBM joined as the CEO and for the first 2 months working, I had Assisto set aside some desks for him and the other 3 or 4 initial employees to work with easy access to Assisto people, build that relationship, and talk to him about the importance it was for me to treat the way he had treated in the past and to continue to let them build their business and we build ours—
Tim: So Assisto was supportive—it sounds like they were supportive and cooperative through this whole process.
Allen: I think they realized that there was nothing they could do to prevent us from creating a subsidiary. And because I had assured them and done everything I could to try to make it a forward step for both of us, rather than taking away their business and replacing it with an organization built from scratch, they were quite mature and cooperative throughout the whole process.
Tim: Do you think that’s a typical result of that?
Allen: I think it’s very unusual.
Tim: Yeah, I do as well.
Allen: What makes it particularly interesting and unusual is that within a couple of months of joining very astutely and correctly realized that because, at the time, Oracle was actually tied for number two, possibly number three, in market share for relational databases in Japan. The reason being that when InformEx, a second tier competitor and Unify, a third tier competitor in the United States, had come in they had just given the source code to them and said, “We don’t want to deal with Japan. We don’t know anything about it,” and InformEx was carried by a company Askey, which along with Softmake, was the premiere personal computer software in the 1980’s in Japan and they did an amazing job localizing the product, selling it, marketing it. So InformEx was far and away the number one relational database product in Japan in 1990, even after we got our business of $10 million in revenues by 1900, InformEx was bigger. And Unify, which was almost a non-player in the United States, had also done a deal with a small distributer that was very committed. But both of them had built very good, solid Japanese language implementations about 3 years before Oracle had finally rolled out our Japanese version. So he realized we’re not number one in Japan. Compared to InformEx and Unify in the United States, we’re huge. It’s not even a question who is dominating the database. Ingres is not here yet and they realized that the only reason we were getting people to take a serious look at our accounting and manufacturing in the United States was because we owned their database. So it was an easy—drop this application from the database vendor on top of the platform, like buying an application from IBM in the old days. The platform had shifted from the hardware to the middleware and we were using their database but we were using their applications too. So he realized that without dominance in database, in the same way that we had in the United States by then, he did not really have a solid footing from which to sell the application.
Tim: So the key was getting dominance in the database—
Allen: He very quickly made the correct decision that what the subsidiary was going to focus on was building the database business, but because he’d gotten to know the folks at Assisto, and realized there was plenty of market for us to address without having to step on them. We basically designed a model where our goal was to capture about 50% of the Oracle business database in Japan through our direct sales efforts and the remaining 50% always, forever through channels. And to always have this two tier, two pronged approach to the market where we were in winning the accounts were strategically critical either because they were large or they were building an application that was going to be rolled out to many locations. And to never lose a critical deal, and when possible, to bring in, by then, Fujitsu, Hitachi, ENC were all partners, was a partner, Assisto, of course was a critical partner, DCL was still a partner. So we had multiple partners that we could bring in to help get leverage in individual deals and in the marketplace, and built a really, really powerful joint go-to-market model, the channels that kept Assisto growing. Assisto continued to be the largest distributer for Oracle, outside of Oracle, for some 10 years after.
Tim: That makes perfect sense from like a marketing and positioning point of view, but usually when these relationships go bad and what makes these transitions difficult, a lot of it is those relationships. It’s a lot of personal focus. Looking back on it, was there one or two things that you guys did right that you can point to and say, “Doing X made this transition so much easier?”
Allen: I think one was the fact that I’d happen to have the insight that I should spend a little bit of time in Japan understanding what was going on, and as a consequence of that, ended up staying for many years, building deep personal relationships with my co-worker, the distributor. So I think the fact that I was one point of continuity in the relationships, through all three transitions. That I think was helpful. And then the fact that not only was I continuous, but for a good part of that, I was physical in Japan, interacting with all of the players regularly.
Tim: So the relationship with you and their trust with you kind of extended to Oracle—
Allen: Helped to keep everything moving smoothly. But the other thing, of course, once we had hired the CEO, it was business to build, fundamentally in the way that he wanted and the way that Oracle wanted. He perfectly free to ignore my advice and cancel the relationship with , or perfectly free to let his sales reps, whenever a deal was discovered, take the deal and close it as an Oracle deal for their commission, rather than letting the partners close it.
Tim: But he maintained that?
Allen: On multiple occasions, I remember him scolding a sales rep for working a deal and presenting a deal internally as it were his own, that he’d heard elsewhere it had actually been curated by one of the hardware manufacturer partners or by Assisto—
Tim: Okay, so he was giving up short term revenue—
Allen: Actively protecting, prioritizing—if a partner had been working the account before we got into it, he would actively make sure that they got the business at the end of the day.
Tim: We talked about this before a bit, about how Japan, especially enterprise sales, is really a channels game. So you’re saying that even well after that, Oracle was doing 50% of its sales in Japan through channels, and 50% direct—
Allen: That was the original goal that we had set out, was to take it from 100% indirect to 50/50. I don’t think it ever got below 80% channel.
Tim: Really? And that’s very unusual for Oracle globally.
Allen: In the initial meeting with and the new international VP from Japan, we had our first planning meeting where we laid out the plan for the first 18 months of the business. They said, “We think we’re going to try and get this at a 50/50 indirect/direct business model.” He said, “That’s fine; I’m not going to question that. But be very careful because Oracle is a very complex product and any time something goes wrong, your partners will likely blame Oracle and you may lose your reputation in the marketplace. Go ahead and do that but be careful.” And I don’t think anywhere else in the world was doing more than 3 or 4% through partners. Most of the business was to partners, in terms of pre-paid licenses or sales rights and things. And not a lot of business through partners. Or it was very, very oriented towards aggressive direct sales and direct market.
Tim: So now that Oracle is back in control of its own destiny in Japan, was the go-to-market similar to the global playbook or was it more like Assisto?
Allen: As we were going through this process, I said, “Finally, I’m going to have somebody who’s an Oracle person and we can finally do the Oracle marketing way in Japan.” I was so excited about that. Coming from engineering, I don’t know why the Oracle marketing method was such a thing I cared about but even with that, in that first year, in that very first 18-month plan, I think we were coming off of 5 and going to 10, was the goal that year. in his budget had put a $1 million marketing budget and as we were talking about what were the plans marketing, this was a couple months after and he had kind of figured out already that dominating the database was the most important thing. He said, “In the spring, after we’ve got a few partners, I’ve gotten to know the lay of the land, we’re going to spend $1 million on a single event in Tokyo. We’re going to invite all of our partners and try and get a couple thousand people to come to this, and do something that looks like we’re so much bigger than, for example, everybody that it looks like it correlates to the size that Oracle is worldwide.”
Tim: So the whole budget is on this one—
Allen: So the whole budget is on one event. It’s one event and part of the budget was $300,000 to take out a full two-page ad in the Nikkei newspaper, once a week for 3 weeks leading up to the event. A lot of efforts to get people in the seats. The way of marketing is make people say wow. Make people, if they’re not doing this, they’re falling behind the movement. So today Oracle open world and dream force are infamous for making San Francisco overcrowded and impossible to get around for two weeks every fall. Well, you can blame and Oracle Japan for that because Oracle open world actually started in Japan. We spent a million dollars, everything on one event, to make people say, “Wow, Oracle is much bigger than I thought.” Larry over for presentation, rented the Royal Park Hotel ballroom, so at least he did it in the ballroom. And we did do direct mail and to get people into the room, so there was a seminar and there was a lot of effort to get people into the room but it was not a regular cadence, you know, 30 people in a room, intimate, and there was none of the telemarketing prequalifications stuff. That never really took off.
Tim: So it was an event.
Allen: It was an event. It was event marketing. And a couple years later, we all went to the Oracle user conference. There was a user conference every year in Oracle and by the second year, I think that , there were about 6,000 people that came to the Oracle user conference and that was—we invited the users and the user presentations, we would talk about what new products are coming that you can expect next year, and it was kind of an intimate thing with the customers. We went to that, and we came home, and and the very ambitious sales VP said, “What can we do in Japan that will be a bigger event than Oracle’s user conference? How in Japan, when we only have a few hundred users, how can we get 6,000 or more people in a room?” So we spent a while brainstorming and the conclusion was, first, we don’t have that many users so it can’t be users-only. So users and prospects, and anybody who wants to learn about Oracle should be invited to come. We’ll open it up to be a marketing event for prospects, much like the very first even that we did. But even that’s not going to be enough, so what we should do is not call it an Oracle event per se, but open systems was beginning to be very much in the press. And Sun Microsystems had really popularized that. So let’s make a statement in Japan that Oracle is not just the relational database, but we are the leader of movement to open systems.
Tim: So kind of co-opt that whole movement?
Allen: Let’s own the open systems movement and invite all the open systems partners, all the software vendors, talk about how open systems are better than proprietary mainframes, and how the world is moving this direction.
Tim: And this approach was unique to Japan?
Allen: This was figuring out, “How can we get 6,000 people in the room?” So it boiled down to a Mark Benioff-like, “How can we do a huge event that raises a lot of attention and people say, ‘Wow, Oracle is so influential and big.’?” So we went to the Yokohama Pacifico convention center, we had contests for our distributers, we had numbered tickets, and whichever distributor gave the most tickets whose customers came to the show would get a free ticket to take to the people at the next year’s Oracle user conference in California. We literally had people passing out tickets to people getting off the train stations in Sakuragicho. “There’s this amazing event! You have to go!”
Tim: Just to get bodies in the seats!
Allen: So we had high school students and mothers coming to it and there were about 10,000 people that showed up to this. We planned for about 6,000. You could not walk through the trade show area, every session was over-booked and standing room only. I think that experience of seeing those crowds, making it impossible to move from booth-to-booth in the trade show area of the convention center, is exactly what Mark is now replicating with Dreamforce in downtown San Francisco every year. So how do we do something that’s bigger than a user conference, rebranding it, renaming it, making it something bigger, and then the standing room only—you can’t walk through the trade show.
Tim: You’ve got to pay attention to this.
Allen: How come there are so many people interested in Oracle? So after that, every year, it was like, “How many more thousand can we get?” I think by the third year, one of the other things that Mark does at Dreamforce and what Open World does in California now, is there’s always a rock concert. So we had to move to Makuhari Messe after a couple of years and said, “Well how can we get 20,000 people to come to this thing?” And says, “I think we need to do a rock concert and just count people who come to the rock concert as part of the attendees.” So we had do a rock concert at the venue on the first night of the two-day event, and counted people who came to the rock concert as part of the attendees to get to, I think, 50,000.
Tim: So is that strategy still kind of the backbone of the marketing?
Allen: Well, it continued for some time after in Japan, and we didn’t continue to try to do everything we could to grow the numbers. I think the last few, there was a deliberate brainstorming, “What crazy new things can we do this year to get more people?” I think the concert was the last year that was a key part of the strategy planning for it. And by then, the business itself had grown enough that we had enough users and prospects. We actually could have 10 or 20,000 people come to an Oracle event just to actually talk about Oracle and learn about the applications and everything that’s going on at Oracle. But Mark, in his informative years and experience, kind of saw that whole evolution in Japan, how to get bigger and bigger every year, and how Oracle had opened up. And then maybe after about the third year in the United States, Oracle stopped doing the user conference and switched to Oracle Open World. You have Yokohama Pacifico and Oracle Open World overcrowding there, depending what happens with Dreamforce and—by the time we did them the second time, got his own way. It’s brilliant. It works in Japan. I finally gave up trying to do the Oracle marketing way because the business was growing well, we were becoming famous and successful in Japan. I think it took us a couple of years to finally catch up and pass InformEx in market share in Japan, but had a very different way of marketing it. Actually, Oracle still does the lead qualification, telesales, telemarketing—they still do the old Oracle marketing way, although you don’t see advertisements to come out to a seminar. But the part of the first 15 years of Oracle market that still exists is the huge Oracle Open World event that actually originated in Japan.
Tim: Excellent. So that’s great. At this point the market entry’s a success and you’ve got a Japanese subsidiary just growing steadily on their own power. Let me ask you kind of looking back on it, sort of a couple of personal questions. What did you think was the most challenging thing for you personally, kind of being stuck as that center point between the Japanese entity and the U.S. entity?
Allen: Except with Mark Benioff, it was always hard to get people in the the product organization to design in the Japanese country requirements in . Sometimes the executives would, in principle, want to do it, and would say the right things and seemed sincere, but when push comes to shove, their development is always delayed, there’s always pressure to roll out some new set of features that a critical customer in the U.S. is demanding. And the Japanese language support was almost never prioritized early in the development cycle. It was always sort of once we finished building these critical new features that Larry has promised the marketplace, yes, we will build Japanese into this. We’ll reintroduce the product, clipping and slipping, and now some customers in the U.S. that are important are beginning to get angry because it’s slipping, and it’s, “I’m sorry, we’re going to have to wait until next week.” So that—
Tim: How’d you solve it? What did you do about that?
Allen: The first thing I did to solve it almost got me fired. When I had been building the Japanese language version in Oracle, I finished the coding in 1987, just as we were having the conversation with Assisto, and it was supposed to be integrated into the Oracle version, 5.0 release of the product and that didn’t happen. I was about 4 months in the company and then after about my year anniversary, 5.1 was going to be produced and it was going to be there. And it didn’t happen again. And it was all teed up, coded, ready to go, and I had built relations with people in development so they knew who I was and there had been conversations to do it. And as we get close to the release date, and I hadn’t been integrated in and they had me wait and wait, and then finally this conversation just mentioned of, “We’ve got to ship it now! Next release!” happened twice in one year. And so part of my decision, when I came to help Assisto in Japan is because part of my job in Oracle international was to provide technical support to the hardware manufacturer OEMs who had received an Oracle porting kit, which essentially was uncommented source code and a little bit of documentation about what you need to tweak to make it work on an iteration of UNIX. So Toshiba had a copy of that in Japan, DCL actually had a copy of that in Japan—that’s how they built their Japanese prototype—ICL and BOL had copies of it in Europe. And I was supporting all these people, and because of my job to get the localization working, I had access to the commented source code. Knowing that there was uncommented source code, called a porting kit, in place in Japan, and knowing that I would be so much more efficient, or anybody would be much more efficient, at coding and continuing in the hands of the Japanese if we had comments. I secretly dumped the entire Oracle code base with comments, brought it to Japan, installed it on a VEX computer, had Assisto buy a MicroVEX, we installed it, and one of their engineers went right to work on integrating all the Japanese into a locally—
Tim: You just did it yourselves?
Allen: They just did it without permission. I took it without permission and at one point, after I had been in Japan 5 or 6 months with Assisto, we pretty much had it ready to work but the engineer who was working on it had, in the meantime, been doing support. We had technical support at questions and interactions with the international team at headquarters and one day he sent a fax—it was still not even—it was faxes—he goes, “I found this comment in the code. What does this mean? What is this doing?” and he asked a question that referenced the comments and that then triggered, “Hmm, there is comments in the source code in Japan,” and of course, John, my boss, immediately knew how it had gotten there, and he called me and said, “Allen, did you take commented source code when you went with you to Japan?” I said, “Yes,” and he said, “You know you’re not supposed to—“ “Yeah, I know.” He said, “Well, this is how I found out and of course I had to tell Larry, and he screamed at me, and technically I should have to fire you, but I understand why you did it. I explained to Larry why you did it and you must never do it again, and no source code imports are ever to go outside of Oracle headquarters without Larry’s personal permission, so if anything like this ever happens again, you will be fired, but I’ve gone to bat for you this one time. You may leave it there now and it must stay there at Assisto, where you are. Don’t ever do anything like that again.” And I knew that if I were to ask Larry for permission to take the VAX to Japan, because the relationship with Assisto was brand new, “It’s the distributor, they haven’t paid us licensing fees for it,” he would have said no. If I hadn’t done it, we would not have had a VAX product to sell for—it wasn’t until version 7, two years into the subsidiary, four years later, before Oracle headquarters finally had an integrated product that went out of the door with Japanese. And in the meantime, there was a team in the international division that was helping with port, we had a team in Korea, and I had a team in Japan who were porting the Japanese version Oracle created in Japan onto all the platforms that Oracle was doing the English porting for. That decision to do something that I, to this day, believe was morally right and necessary, and in terms of this assessment, correct and necessary, but legally illegal, and potentially could have cost me my job.
Tim: You can get away with a move like that once, maybe.
Allen: You can’t. And I’m sure if it had not been John Luongo there going to bad for me—if someone else reported, if I had not had—if I hadn’t spent a year with John saying, “I’m ready, but I can’t get it rolled in,” he knew that I had been ready to build this out for a while. My relationship with John, and he had seen what I’d been doing up to that time to make it happen the proper way, that I think made him willing to go to bat for me, but he understood the rationale, he understood me as a person, we by then had gotten to know the Assisto as a team well enough, and I had physically located myself within Assisto, that the whole collection of things made him willing to go to bat for me. But, subsequent to that, I was not able to get access to all the applications code for example or other products. So after that, it became sort of an ongoing process of trying to convince—and Mark Benioff, to his credit, it was the only product executive in my history with Oracle who, from the beginning of designing a product, he brought me into his office one day and said, “I have this really cool object-oriented tool that I’m building right now. You should see it,” and he had his engineer show it to me. He says, “That’s really cool! You know what else would be cool? If this was the first product that ever goes out the door from Oracle with Japanese language built in. It’s object-oriented; it should be able to do it!” And he said, “Yeah, yeah, do it. Sit down with Roberts and tell him what he needs to do.” So I sat down with his engineer and got it working on a Japanese PC, said, “Here’s where it breaks and we’ll send an engineer over to help you do it. Keep trying and working it.” And ultimately, the way Oracle ended up solving the problem is we sent a fairly large contingent of engineers—I think ultimately it was about 70 people, once the Japanese company was 6 or 700. We had a group of Oracle Japan employees in technical support, in different development organizations, all of whom were tasked with wither high priority bug-fixing for critical customers or ongoing maintenance and enhancement of Japanese language features in different products. We sent Japanese engineers, paid for them, sent them to headquarters, and the development at headquarters were willing to integrate these Japanese engineers into their development organizations.
Tim: So obviously you don’t recommend that move, but—
Allen: In that case, I didn’t see any other way to avoid, one year later, having the same conversations and attitudes, “We don’t have a Japanese product. We can’t sell without a Japanese product.” Because that excuse, that made sense to me that that would be, especially in a world where Unify and InformEx were both in the market with good Japanese products, to not have a Japanese product, it was clear to me that was a problem that had to be solved.
Tim: What advice would you give to a new country manager or someone tasked with doing this job? With bringing a new technology or product into Japan?
Allen: With the joint ventures we’ve done subsequent to my experience at Oracle, with Salesforce, with Concur, with Marketo—I guess not with Concur—we have found it very useful to have someone at headquarters be in Japan for a few years as sort of the advocate for the president’s agenda back to headquarters.
Tim: So an advocate for the Japan team?
Allen: An advocate for the Japan team back to headquarters. In ever organizational scene, there are plenty of people at headquarters who will tell Japan the way headquarters does things, and will encourage them to do things the headquarters’ way.
Tim: It would seem to me that would be the default mind-set of someone coming over from headquarters.
Allen: It was mine, “The Oracle marketing way is the way we should do it!” There are plenty of people who will fights those battles. One of the things I remember being frustrated in the early days at Oracle, there were opportunities that we were working on with Assisto where there were maybe 40 or 50 potential Oracle licenses to be sold if we would port to the SGI workstation—that was one I very specifically remember—but we didn’t have Oracle working on the SGI workstation. And the conversation was all, “If there are potentially 50 licenses after this initial test license to build the application for team deployment, why can’t we get Hitachi to buy 50 licenses right now upfront?” And then we’ll build the dailies for them. The fact that Unify was already there in Japanese, and was fast enough to do what they wanted to do, somehow wasn’t persuasive enough to get—we actually lost a significant volume of business to Hitachi because we couldn’t get our heads around making the effort to build one port of a product for one type systems and believe that, in fact, Hitachi would actually follow through and buy the whole 50 systems they ultimately expected.
Tim: That makes sense though.
Allen: Even though there was a competitive product that they could roll out, even though they would prefer to use Oracle, they had been convinced Oracle would be the better product, they would rather work with us, we weren’t willing to make that effort and investment to do that.
Tim: But I could see if that person is Japan’s advocate to headquarters, it makes sense. There will be a lot more trust from the Japan organization, the trust of headquarters, and they’re being heard.
Allen: The key thing that I tell these people, when we’re bringing them to Japan, or I’m doing my initial coaching and setting the goals and parameters with them, is that, what I’ve just said to you, there are plenty of people who are going to tell Japan how headquarters does things and helps them there. I said there will be plenty of times when the CEO is telling you he needs this, or that special treatment or discount, to make that deal happen, and you also personally will question whether that’s really the case. And whether, really, if headquarters does that, can he really close the business? You’ll find sometimes you even question it. And if you have serious doubts and you can tease him out and help him come to his own conclusion that it’s not going to happen, great. But otherwise, if you will go to bat for him, every time he has a weird request, whether you fully understand it and fully believe it or not, once you’ve established over a year or so a pattern of, “I’ve got your back,” I said to give him two strikes. Go to bat for him. If it doesn’t work out and people at headquarters are complaining, you promised this it was going to work out and it didn’t, it doesn’t always work out. Give him two strikes and until there are three times that he makes you jump through hoops to get a deal done that he can’t close, assume that he’s telling the truth in what he needs back. And after the third time, maybe have a serious conversation with him. By that time, once you’ve gone to bat a couple of times and done that for him, and defended him even after the deal didn’t close, then when there is something that headquarters believes is the right way to do it, and you believe is the right way to do it, and the CEO does not, then you’ll have much more credibility explaining to him this is something that is not debatable with headquarters because of the president’s attitudes, or it is so built into the culture at headquarters—it’s not the way it’s done in Japan but it’s not a negotiable thing. Because you’ve gone to bat for him, he knows that you already have his back and that even telling him that you have to do this, it’s because you have his back and you think it’s important for him to be successful and trusted by headquarters, that you then have a different context with which to do it. So I think starting by making it clear to him that you’re on his side, through multiple experiences, and then when there are things that really are important for the success of the business in Japan, for that person’s success as a leader, and that really matter to critical people at headquarters, then you’ll have a much higher likelihood of getting him to listen to you.
Tim: Excellent. So before we wrap up, is there anything you want to add or anything I should have asked you that I forgot to?
Allen: I think in thinking back on Oracle’s history, and all the other companies that I’ve worked for, the ones that have been most successful, there’s one common thread, and that is that the country managers in Japan, every single quarter of their careers with their company deliver the results that they promise at the beginning of the year. The headquarters’ counterpart to that is that in every one of those successful companies, headquarters has not tried to control how those results are delivered. Headquarters has been extremely flexible and open-minded in allowing the CEO to do it the way that he believes it should be done in Japan. So there’s a combination of tremendous flexibility and support for the CEOs’ innovations and creativity, and what’s comfortable for that person that make it possible for him to deliver the results. But there is the corresponding with that flexibility, taking the responsibility to always deliver what was promised at the beginning of the year. And that creates sort of a long-term trust and opportunity for innovation. And that flexibility combined with committed capable leadership has been the common thread throughout entire career and with each of the most successful companies we’ve worked with since has been that.
Tim: Wow. Thanks so much for spending this time. There’s an amazing amount of material in here.
Allen: You’re welcome.
And we’re back.
What I found most interesting but perhaps least surprising, about Oracle’s market entry was the importance of letting the Japanese office develop and run their own marketing and sales strategies. Now, this is hard to do and it requires a great deal of trust in the local team. I mean, you’ve got an optimized strategy with a proven track record, one that you know will move product. It’s both logically and emotionally difficult to let Japan try something new and, in fact, it usually is best to start with a global playbook unless there’s a very good reason not to. Now another thing Oracle had going for it, although we kind of glossed over this in the interview, is that the Japan head was reporting into the head of international and not the senior vice president of global sales. Great sales managers are often execution machines who have little time for excuses or claims that the sales process needs to be customized. We’ll be talking a lot more about this in future episodes.
You know, it’s kind of amazing. Oracle’s market entry happened more than 25 years ago but the challenges they faced and the way they overcame them provide a perfect blueprint for the companies who are entering the market today. In some ways, it’s surprising how little changes in the technology industry.
If you’ve got questions about Oracle’s market entry, or you want to share some thoughts on our new expanded format here at Disrupting Japan, Allen and I would love to hear from you. So come by to DisruptingJapan.com/show058 and let’s talk about it. When you drop by, you’ll find all the links and sites that Allen and I talked about and much, much more in the resources section of the post. But most of all, thanks for listening and thank you for letting people interested in Japanese start-ups know about the show. I’m Tim Romero and thanks for listening to Disrupting Japan.
But most of all, thanks for listening and thank you for letting people interested in Japanese start-ups know about the show. I’m Tim Romero and thanks for listening to Disrupting Japan.
I’m Tim Romero and thanks for listening to Disrupting Japan.