April 7, 2005 — Market analysts sometimes get a bad rap — for being overly optimistic, overemphasizing a trend or being just plain wrong. But when the data gathering’s done and the modeling completed, it’s still as much an art as a science. Small Times’ David Forman turned the tables on four analysts, putting them in the position of answering rather than asking the questions.
They are Thomas Abraham, vice president of research at Business Communications Co., Kimberly Allen, director of display technology and strategy at iSuppli Corp., Atakan Ozbek, principal analyst at ABI Research and Jim Walker, vice president of research for semiconductor and emerging technology at Gartner Dataquest. Their responses show that it’s a difficult game to play, especially in emerging technologies.
What are the toughest challenges you face in performing market analysis?
ABRAHAM: Since many of the new topics on which we conduct market analyses are new technologies and new products, the companies are very reluctant to speak about them. Therefore conducting successful primary interviews is a challenging task. It would help if you have prior contacts in the company. We have a dozen newsletters on new technologies and we also organize conferences in some of these technologies. These efforts put us in touch with the companies on a regular basis.
Once one has all the feedback on the markets and industry structure, the next big challenge is to make reasonable estimates of the current and future markets. Since different people offer differing numbers, one has to compile and make the best estimates.
ALLEN: Market research, like all research, is the synthesis of many disparate pieces of information into a coherent story. It is difficult enough to know when there are enough pieces to form a reliable narrative, but in market research there is also quite a bit of doubt about the accuracy of each piece. If I speak with 75 percent of the companies in a certain area, is that enough? What if half of those were stretching the truth in significant ways?
OZBEK: While collecting intelligence that encompasses reality and market potential is not easy, analyzing the intelligence into a format that will prove valuable to a client is the challenging part.
WALKER: There are two of them. One is making sure that the data is accurate. Two is making sure that you’re talking to the right person. A lot of the time you get a name of a contact but that may not be the right person. It takes a lot of time. You ask them a little bit about their background, some of their publications, the technical societies. Then you ask them to give you some market numbers. If the numbers are accurate, you know they’re well informed.
How is covering emerging technology markets different from covering established, mature industries?
ABRAHAM: For emerging markets, there are no published data. Established industries generally have government data or there are trade organizations which compile such data. However, such published data need not necessarily be accurate since they reflect only data from those who participate in such surveys. That is where market research companies play their role in looking at the published data for their accuracies and filling in the gaps in the data. For emerging technologies, one needs to talk to players and get primary information.
ALLEN: The largest difference is that everything is a zero-billion-dollar market. Revenue barely exists, but it has the potential to be quite large if a number of things come together. The trick for the analyst lies in seeing all the pieces that need to be in place for the market to grow, as well as the context of those pieces in the already existing industry. New technologies do not emerge into a vacuum; they emerge into the real world, even if they are serving entirely new applications. Handling new technologies is thus much more of an art form than for established technologies.
A second difference is that small changes greatly affect the total market. If there are only four products on the market containing a new technology, a delayed shipment can result in decreased revenue for the quarter in the tens of percent. Issues like this are lost in the noise of a huge market with dozens of products.
OZBEK: Since there is very limited tangible information available, it becomes essential to understand technological fundamentals. However, technological fundamentals will not be sufficient if the companies or emerging technologies do not provide superior value to end users. Market research and projections will have to reflect these underlying market dynamics of the emerging sector.
WALKER: There’s no concrete data available. You have to develop methodology. It’s really a case-by-case basis. It depends on what type of emerging technology you’re talking about. Sometimes there’s government data. A lot of times the way you do it is you develop a survey and they give you feedback that gives you a baseline.
You look at the manufacturing process. Based on that you call some of the equipment companies. You develop a model based on a manufacturing process. You can quantify it. For example, I cover intimately the IC packaging area. Flip chip has been around for years but hadn’t been used in corporate products because it was too expensive. Then a market driver came out — the cell phone — and it became a consumer product. That whole industry changed.
Is there such a thing as a nanotech market? A “small tech” market? Is it useful to define such technology areas as markets? If so, how do you define the space that has evolved at the convergence of micro and nanotechnology?
ABRAHAM: There is a nanotech market with several nano segments. For the purposes of definition, we call “nano” all those materials which are below 100 nanometer and those technologies which use nanomaterials in their components and devices. We also lump all those technologies which use atomic scale processing to evolve a product as nano. “Small tech” or micro technology did not establish as a catchy title, although micro devices such as MEMS are increasingly becoming established markets. Since the boundary between micro and nano is narrow, some people also lump micro into nanotech markets.
ALLEN: The definition of “nanotech” that ties it to some length scale will not ultimately be useful. After all, there is no distinct market for objects that are around one meter in size, or one centimeter. Although markets always have hazy boundaries, they do define some relatively coherent set of products and services that affect each other. Technologies spanning medicine, biotech, industrial materials, automotive and consumer products do not represent a distinct “market.”
The real issue here is that developers of nanotech look out upon the existing market landscape and see different boundaries between fields. It’s like looking at a map and wanting to redraw the borders (or erase some) based on this new technology. It’s natural to express this wish by articulating a new vision for the market boundaries.
What will happen in reality is that the nanotech vision will interact with the current state of affairs to produce something different from either picture. And it will continue to evolve for a long time, in a sort of dialogue.
OZBEK: Yes and no. Yes, we can define it as a market due to strong continuous interest by a wide range of companies in the field. No, there is still no marketable commercial product. It will take a little longer for the market to deliver reliable product to end users.
WALKER: This is the whole area of hype. The reason I say that is that nanotech is probably one of the most misused words. In my mind, nanotechnology is an additive manufacturing process on the nanoscale. That’s where nanotechnology differs from the others.
What sort of methodologies do you employ? Who do you interview or survey for your research?
ABRAHAM: To obtain primary information, we use telephone interviews as well as questionnaires sent as e-mails. The people interviewed are generally the director or vice president of technology, marketing, sales or project managers, or the presidents of startup companies.
ALLEN: The key for me is diversity. I speak with the supply side and the demand side, the engineers and the executives, the big firms and the startups. I am often delighted by the wide range of views I encounter regarding a technology. Other times, everyone seems to be repeating the same party line. I instinctively doubt party lines.
OZBEK: We employ four different tiers of methodology: in-house analyses, primary and secondary research, judgment sampling and feedback.
We speak to executives, technology and business management personnel, competitors, government bodies, and about any stakeholder that could provide a true understanding of market dynamics, which will help us dissect the industry being assessed.
WALKER: We do much more data gathering and data analysis to support the qualitative hypothesis and research that Gartner corporate comes up with. There’s different ways. You do what’s called bottom-up and top-down. If they correlate within 10 percent you’re doing really well.
If there was one thing that could make your job easier and your forecasts better, what would it be?
ABRAHAM: If people being interviewed are willing to talk to the market researchers as well as provide good info.
ALLEN: A better memory!
OZBEK: Limiting NDAs (non-disclosure agreements) would make our jobs easier since it will allow companies to talk about their alliances. This will also allow market analysts to make more accurate competitor comparisons.
WALKER: Have everybody be open and give you their proprietary information. Naturally, a lot of times they don’t want to.
How do you define success with a forecast even if it’s not 100 percent accurate? How do you know what you’re being told is true?
ABRAHAM: There is no such thing as 100 percent accuracy. When we forecast, that is valid to the date when the forecast is made. From the forecast date to the future dates, anything can happen. However, when we do the forecast, we make assumptions and provide realistic feedback from the industry on the likely scenario of the future markets. From that we evolve a reasonable accuracy on the future markets. We can call our forecast a success if it is within 20 percent of the projected future market numbers.
ALLEN: Given that there are infinitely more ways to be wrong than right, it’s important not to define success as 100 percent accuracy. Instead, I focus on stating what I believe to be accurate at any given time, but am always willing to revise in light of new data. Besides, if my forecasts were perfect, clients wouldn’t need to renew very often.
OZBEK: Forecasting is imperative but it is only one part of market research. Providing value means understanding the core value of technology, management’s business insights and putting all of the pertinent information in a context that reflects the trends and needs of the industry being assessed.
WALKER: You don’t (know if it’s true). You do a lot of surveys to validate the data. But I think people are realizing “garbage in, garbage out.” They realize if they give a bad number and four others give a bad number then it has a cumulative effect.