Here’s a continuation of the series of interviews designed to explore the landscape of companies creating business simulations and other innovative experiential learning tools. I first became aware of Advanced Competitive Strategies (ACS) when I read the article Game Over in FastCompany magazine. The article, subtitled “Now let’s get down to business. Simulation lets you play to learn – without having to play for keeps”, focuses on competitive strategy simulations. It details certain companies that have “borrowed many techniques from the Pentagon’s experience with computer-based war games” to create a new breed of “business war games”:
One of those quoted is ACS founder and CEO Mark Chussil. I’ve since had a few communication exchanges with Mark and was excited to ask him some more questions about how he uses simulations in his business.
How long has ACS been in business and how did the company begin?
Even in the best of times business strategy makes chess look simple and the tools strategists have had to work with are not up to the task. We wanted to create a powerful, realistic way for strategists to explore their high-impact decisions before committing to a path.
ACS has been running business war games and strategy simulations since 1992 and we began work on our core technology around 1986. We were excited at the ways we could combine computer technology (which was just getting going at the time) with principles of competitive strategy to provide seriously new forms of strategy analysis that would help strategists achieve competitive advantage. The combination of computer technology and strategy models meant that all of a sudden we could do things that no one could have done before.
You mention business war games and strategy simulations separately. What’s the difference?
A business war game is an interactive process involving teams of people role-playing their own business, competitors’ businesses, customers, and other entities relevant to the problem at hand. It may include a computer simulator to calculate the performance results of the teams’ moves, in which case we call it a quantitative war game, or it may use human judges or umpires to estimate the goodness of the teams’ moves, in which case it’s a qualitative war game.
A strategy simulator is a computer program that takes available data and estimates about a market, combines it with business moves and customer reactions, and calculates outcomes such as profitability and market share. Strategy problems are different from accounting problems, which is why ACS doesn’t use budget-style spreadsheets. Instead, we’ve built a series of simulators that use strategy-related factors (product quality, customer loyalty, price sensitivities, cost structures, response times, and so on).
So, a business war game may or may not use a computer-based strategy simulator, and a strategy simulator may be used with or without a business war game.
You describe “Business War Games” as “the strategy equivalent of clinical trials, flight simulators, and sparring partners, all together.” Can you elaborate on this – what is it about business war games that makes them so compelling?
Pharmaceutical companies use clinical trials to know whether a drug works or doesn’t; in effect, they’re doing rigorous what-if tests. Business war games and strategy simulations, especially the decision-tournament kind of simulation, are the business equivalent of those what-if tests. They go way beyond best-case, worst-case, likely-case sensitivity analysis.
Think about what you do with a flight simulator. You don’t turn on the autopilot, sit back, and watch the plane cruise across the Pacific for 10 hours. No, you try to handle the plane in a bad storm, you try to land after dark when you’re running out of fuel, all the tough stuff that’s too dangerous to do in real life but that you want to be prepared for. The worst thing that happens in the flight simulator is that you hear an electronic crunch and you press the reset button. It works the same way in a business war game: if your strategy doesn’t pass the stress-test, you press the reset button and try something else.
Boxers train with sparring partners because that’s how they build skill. It’s much better to train and even lose with a tough sparring partner than to beat a wimpy sparring partner and lose when you’re up against a real opponent. Frankly, the spreadsheets that most companies use are easy sparring partners; they don’t fight back! In effect, spreadsheets bake in the assumption that your strategy will work. Business war games, on the other hand, build in competition: they have your own people figuring out the best way to attack your business. It’s much better to face those attacks in a war game than to experience them for the first time after you’ve risked careers and big bucks on a strategy.
I might have a special appreciation and affection for simulators. I was on an airplane a few years ago and an engine blew up and burned in mid-air. The pilot was well-prepared and knew what to do, even though odds are that he hadn’t previously dealt with that kind of emergency other than in a simulator. Not to mention the aircraft designers, who must have tested their planes very thoroughly indeed.
What role do business war games and strategy simulations play in higher-level strategy?
Great question. At one level, we could say that business war games and strategy simulations are another tool in the strategist’s toolbox; that is, they’re a way to develop and test strategy options. That’s a technological perspective: we look at the costs and benefits of war games and simulations, and we compare them to the traditional processes companies use. I think it’s a slam dunk for the appropriate problems — strategizing under uncertainty, new-product introductions, defending against competitive moves, evaluating strategy options — but I appreciate that the techniques are unfamiliar and can be technically confusing.
In another sense, business war games and strategy simulations are at a very high level because they are all about how we think and how we lead. I don’t care what strategy comes out of the war games or simulations I conduct; I tell my clients that I don’t care if they zig or zag. What I care about, and I care about this a lot, is the quality of their strategy decisions. That’s what these techniques are designed to improve. No decision comes with a guarantee of success, but the better the decision the higher the odds of success.
War games and simulations almost always lead to surprises. Why? Not because they have more decimal points or curvier demand curves. Rather, because they help us think better. They help us take competitors into account realistically, not optimistically. They help us focus on cause and effect. They help us think things through and see the big picture. They help us learn to think strategically. (One of my clients, a very thoughtful and very senior executive, used business war games as much to groom executives as to make strategy decisions.) Isn’t that what top management wants in strategy development? Isn’t that the direction in which top management wants to lead?
I’d love to get a sense of a few of the business simulations you’ve been involved with. Tell us about ValueWar™ and Top Pricer™.
There’s something very cool and very important about the simulations my colleagues and I have built. I remember whenan early version of ValueWar generated some numbers that surprised me. I traced the calculations through the program, and everything looked right. Then it struck me: the simulation was right, and my expectations were wrong. When I made that leap, I was able to get new insight from the model. I was learning from the model. And that sense of discovery has happened in every business war game I’ve conducted. Not just for me, but for the strategists participating in the game.
ValueWar is ACS’ core technology. We spent 17 years figuring it out and field-testing it with our clients. It can take into account a couple hundred strategically relevant factors, and it won an industry award a few years ago. It’s basically a large collection of simple, non-controversial concepts that describe the competitors, customers, and events that describe a market. How does it work to create such a model? For instance, every market has a growth rate, so we can write growth-rate equations in advance without knowing the specific growth rate to simulate. The actual growth-rate numbers are part of custom-calibrating the model for a specific situation. In most of our client work, we use subsets of ValueWar that include the factors important to the problem they want to address.
The Top Pricer simulator is an example of what ACS calls a decision tournament™. In effect it collects pricing decisions from strategists and plays them in all possible three-competitor combinations. Yes, that means it can run millions of simulations. It lets us see what might happen to a given strategy under a very wide range of possible scenarios, which means that we can assess the robustness of a strategy. In other words, we can see how vulnerable a strategy is to competitors’ moves or environmental changes. That simulator is set up for pricing decisions, but it could easily be adapted to many other decisions.
I designed the DXMA™ simulator for Crisis Simulations International, a company in which I am a founder. We expect it to be awarded a patent. It’s a system that lets a program designer put together a highly interactive, real-time experience for multiple decision-makers. It can be pretty intense: CSI simulated a terrorist bomb going off in an American city, and I role-played the mayor in a test of the simulation, and after an hour I wanted to run from the room. I have huge respect for the public officials who step up to the responsibility of making excruciating decisions under enormous pressure with imperfect and even contradictory information.
We’ve built other simulations, and there are war stories for each one, and it’s probably time for me to stop talking about them!
One thing you and I had discussed was the “Decision Tournament” nature of Top Pricer. Whereas many simulations (Harvard Business Publishing’s included) create “worlds” of users mapped to a given class or company program, your simulations assess the performance of any given user against some whole or partial performance snapshot of all other simulation users to date. These “collective results” allow you to in essence have other users inform the competitive dynamics within which any single user operates. What drove you to this type of approach and why do you take the extra time and effort to design Top Pricer this way?
The Top Pricer decision tournament works as you described because I wanted to produce a truly powerful what-if machine. There’s a lot we can do with today’s computer hardware that would have been impracticable or unpleasant even a few years ago.
Even a conceptually small pricing problem — three competitors in a market, twelve quarters of pricing decisions — has an enormous number of permutations. Imagine that each competitor can cut, hold, or raise its price in each quarter. For a single competitor, that’s 3 to the 12th power, or 531,441 possible moves. For three competitors, that’s 531,441 to the 3rd power, or roughly 150 quadrillion. On my computer, running 50,000 simulations a second, all those permutations would take 95,124 years to run, which is beyond most companies’ planning horizons. A two-year, eight-quarter problem would take 65 days, which is still not fun, especially if the computer wants to reboot for a system update after 64 days.
Of course most of those 150 quadrillion simulations are useless. We don’t really care about the difference between “hold price every quarter except raise in the 10th” and “hold price every quarter except raise in the 11th”. The problem is picking the useful simulations. There are other interesting problems that the decision-tournament technique solves — for instance, we shouldn’t assume that competitors behave completely independently, which could have one business blithely raising prices while its competitors keep cutting — but it’s enough to deal just with that one for now.
The decision tournament technology saves strategies from all the players so far, and as new strategies are contributed the decision tournament gets bigger and more realistic. (And by the way, it does not ask players to pick strategies such as “hold cut hold hold raise hold”. It uses a more-sensible technique.) It’s bigger because there are more strategy combinations to test. (It’s in the millions now, which takes about a minute.) It’s more realistic because, like a poll or survey, the collection of players’ strategies gets more and more representative.
The Top Pricer tournament is about pricing, but there’s nothing about the decision-tournament technology that restricts it to pricing decisions.
We both have a vested interest in understanding how simulations operate and how their unpredictable nature can be a source of learning even for simulation designers. We’ve had simulation authors who really get this (see my blog entry Teaching with Online Business Simulations) and you wrote about this dynamic yourself (see ACS blog entry When I Was Wrong). Do simulations teach you as much as they teach your customers?
Absolutely. I’m constantly learning from my simulations. As you suggest, a key part of that is from a willingness to be wrong.
I learn from the numbers coming out of the simulations, and from the people making decisions for the simulations.
Here’s an example of what I’ve learned about the numbers. After more than 30 years working with computer-based models, I think I’ve learned a meta-lesson about the numbers. When I see the numbers, I do two things. One is that I look at the numbers. Are these numbers high or low, who’s making what moves, who’s winning. The other is that I ask why the numbers came out as they did. Did we lose market share even though our product quality went up and our price went down? If so, that points me toward competitors’ moves; presumably they raised their quality faster than we did, or cut their prices more than we did, or even both. If we gained share, though, I look at our profits to see if our share went up because we bought share. I also look at competitors’ results to see if we’ve hurt them, because if we did, they’re likely to fight back hard. Remember, you’re competing against human strategists, not against emotionless numbers.
At a deeper level, numbers tell me something about what people are thinking. Why did someone make those moves? That person believes the moves to be good and effective, otherwise he or she would have chosen something different. What’s the rationale implicit in those decisions? I don’t mean that their decisions will work, or that they won’t. I mean only that this smart, experienced, motivated person is trying to succeed, and she or he made decisions accordingly. Or I reason backward from success. In the Top Pricer tournament, I realized what I was doing wrong when I contrasted my strategies with those of the person who has earned the highest score so far. I saw what that person was doing, and it was fundamentally different from, and more effective than, what I’d thought.
Here’s something I’ve learned about people. When ACS runs business war games we always have at least two “rounds;” that is, we run through the simulation once, debrief, then turn back the clock and run it again so people can do something different with their new-found knowledge. It’s a good thing, because the first round is almost always thrown away. In the first round the home team (the client) makes an enthusiastic effort to do out-of-the-box strategizing, to be creative and bold and innovative, and they find themselves shocked at the strength and effectiveness of their competitors’ moves. (Remember, those competitors are being role-played by the client’s own people!) It’s a sobering experience, and extremely useful, because they truly get that their competitors are as dedicated to success as they are. In the second round, the home team always thinks more creatively, and they do much better. So, what I’ve learned is that people have to go through the first round to get to the second. I can tell them in advance what’s going to happen, and it makes no difference. We humans just have to experience it. Luckily, we can get that first-round shock — and figure out the second-round comeback — in the war game.
I learned about people also from this war game. We did a quantitative war game for a company in the #2 position in the market. The war game showed that their chances of overtaking the leader were pretty much zero. The president of the company delivered an impromptu speech about how #2 is the worst place to be. We asked if, given the war-game results, he thought they could be #1. He agreed that wasn’t going to happen. We then asked if he would prefer to lose some share and become #3. He thought about it and said no. The lesson I learned from that is that we humans find it easy to latch onto catchy advice nuggets, things that I call diet-book strategies, but it takes something as rigorous and thought-provoking as a business war game to really figure out what works and what you want.
In yet another simulation, this one focused on marketing, we told teams what their marketing budgets were. We also told them they were free to spend more or less than those budgets, with no limits. When we looked at their decisions, we saw that every decision made by every team was within 5% of their stated budgets. I see that as budget mentality laid bare: don’t overspend the budget because you’ll be in trouble if profits fall, and don’t underspend the budget because you won’t get as much next year. Of course strategists may choose to hug their budgets; the war game added value by making that choice explicit and open to discussion.
Notice that those insights and observations would be hard to get without realistic numbers (if the numbers weren’t there, it’d just be a debate), and that the net effect is all about better thinking and better decisions.
Do you have a favorite recent story of how a simulation you developed really helped a customer?
Definitely. Sorry, though, that I’m not going to name names; strategy development is highly confidential to companies.
We recently did a quantitative war game with a company planning to enter a consumer package goods market against an incumbent with a highly dominant market share. The home team, role-playing themselves, figured that it was a classic introduce-a-new-product exercise, so they adopted a strategy with a heavy dollop of marketing spending aimed at inducing trial… which, given the competitor’s dominance, meant inducing trial among the competitor’s customers. The strategists role-playing the competitor said, in effect, okay, you want to spend money on marketing, we can do that too. So they raised their marketing spending to hold the home team at bay. Dissatisfied with their progress and determined to make real inroads, the home team raised their marketing. Annoyed at the pesky little intruder, the competitor team raised their marketing spending again. At the end of the first round, both teams surveyed the damage and both teams were shocked at how much they’d hurt their profits and at how easily they’d fallen reflexively and emotionally into a marketing war.
In round two the home team thought hard about their assumptions. They realized that they had been positioning themselves as a new, improved alternative to the incumbent. They came up with the idea of reframing the market, of positioning their product as a leap forward. They noticed even that the incumbent’s name could be turned into a liability. Round two of the war game was very different.
Why didn’t the company come up with that strategy at the start?
When they hear me tell these stories, people ask why the war game was necessary. In the war game I just described, and in pretty much every other war game I’ve conducted, there is always a lesson and the lesson always looks obvious… after the war game. The lessons are never obvious before the war game. That’s true even though companies have smart, dedicated, motivated, experienced, well-equipped strategists. The lessons are obvious only after the war game.
If someone is interested in adopting business simulations to enhance learning, what advice would you give them as they explore product options?
Call ACS first. Seriously, though, for business war games, start by taking a look at an article on ACS’ website, Learning Faster Than the Competition. It describes key decisions you’ll have to make as you design your business war game.
For strategy simulations, and also for business war games, look at You’ve Got the Data. Now What? (PDF). It talks about strategic thinking in ways that will help you design or assess simulation models, as well as use them effectively. It’s a trickier subject than business war games, though, because it is both technical and conceptual. Technical because of the calculations involved; conceptual because the way the simulator thinks will determine the insights you can get from it. For example, if you use an accounting-based spreadsheet as the simulator you will get accounting-based analysis, which means you won’t see much about customers or competitors. If you use a forecasting model as the simulator, you will probably see curves and R-squares and so-called optimizations, but if the forecasts are based on history then the model may be seriously misleading if the future doesn’t look like the past. Put another way, extrapolating from the past works well on the easy problems — i.e., when the future resembles the past — and works badly on the hard problems, when you need help the most.
Don’t worry much about precision. No simulation model is “accurate” when it comes to the future, and the purpose of the simulation is to help you make decisions, not to forecast decimals. For more on that subject, see Precision In, Garbage Out.
Rather than precision, focus on realism. Ask how, if at all, does the simulation take into account competitor moves, shifting customer preferences, the link between product quality (as perceived by the customer) and the bottom line, customer loyalty, inertia, or switching costs, the mix of fixed and variable costs, and so on.
Ask whether the simulation model contains magic numbers, which do things like saying that spending $X on marketing leads to Y% market share. Such numbers — benchmarks, historical experience, market research, and so on — may be wrong in times of change. Moreover, they may be hidden or locked, which makes it hard to question assumptions and run what-if tests.
What's on the horizon for ACS?
The core of ACS is in helping strategists tap a deep understanding of competitive strategy through the technology of simulations and war games, whether in corporate headquarters or the classroom. We’re able to do things today that I could barely have imagined five years ago, and I think that five years from now I’ll be as amazed by the next five years’ progress as I am now with the last first years’ progress. Keep watching!
For more information:
Advanced Competitive Strategies
http://whatifyourstrategy.com
Also see this great interview that Mark Chussil did with the Oregon Chapter of the Society of Competitive Intelligence Professionals.
See all the business simulation interviews here:
http://saulnier.typepad.com/learning_technology/2009/06/business-simulation-interview-series.html
Posted by: Denis | August 21, 2009 at 04:31 PM