I’ve spent a lot of time in my career ether looking for, or being asked for market size information. It’s not surprising, as any company that makes new products or enters new markets wants to know what to expect. In my current job, about 9 out of 10 calls I get are folks asking for data on market size, for one market or another.
The problem is that in most markets and for most products such information is hard to obtain. Yes, there are products like cars, airplanes, battleships and houses that are tracked fairly well and have accurate data availble. But the oceans of smaller products, like the number of mailboxes, or tractor, or replacement steering pumps usually lack published data.
Sometimes it’s hard to find just because it’s not published. The total number of excavators sold per year, for example, isn’t really published anywhere publicly.
Sometimes it’s because it’s hard to define. How many economic reports are sold? How many pieces of living room furniture?
Sometimes it’s because so few companies make the item that publishing anything would cause disclosure issues.
While many times there is no replacement for the exact number or value of items sold, I’ve found that most of the time people go looking for more information than they really need.
Usually the information can be obtained, either from a specialty publisher for a fee, or by hiring someone like me to dig it up. But before you commit yourself to needing expensive data, ask yourself the following:
Do you really need to know the size, or just whether or not it’s ‘big enough’?
When folks ask for information I always interrogate them a bit to find out why they need it, and what they’re going to do with it. Very often the answers point to something like ‘Well, we’re thinking about going into this new market, but we need to sell X,000 products to break even, so we’re looking for the size of the market.’ It happens all the time. One question – whether or not we can sell X,000 products – turns into another – how many are sold?
The problem is that these are two different questions. The first is usually pretty cheap to answer. The second is usually expensive.
Many times a really good gut check can be to ask folks connected to the industry whether they think a newcomer could sell x,000 in their first year. The reactions to that question will not only point to feasibility, but also to potential roadblocks.
Many sources of data, such as export records, Department of Commerce, and Statistical Abstracts of the Census will give sales levels for past years that can be used to answer these questions.
What would you do if you had the answer?
This trick is hard to play on yourself, but it’s useful with others. Suppose someone asks how many sewing machines are sold per year without the accessory table. You say ‘suppose it’s 2.6 million – what then?’ They think for a minute, look at their spreadsheet and then ask how many of the people who bought those would buy an accessory table?
When people go on an info hunt they’re often resitant to hearing that they’re looking for the wrong data. It may sound silly, but the simple exercise of giving them some example data and asking what next will often prompt them to understand the gaps in their thinking.
Do you really need to know market size, or the number of opportunities?
Is it really the number of kids who like ice cream, or the number of kids we can reach via our marketing that matters? Response rates for various activities, like direct mail, surveys, etc. are published and can be used to create models in these instances. Your own company likely has enough data to estimate conversion rates as well.
Can you estimate by counting hosts?
Instead of counting windows, how about counting houses? Instead of estimating the number of road signs sold, how about the miles of highway that are constructed?
Methods like this usually don’t yield precision, but they can usually be based on more available data because they are more broad. The nice thing about them is that by building the models that make the data useful, you gain a lot of insight on how the business will operate.
Are you chasing too much precision?
Market size estimations are imprecise. They will never be very accurate, and it doesn’t really matter because business plans are also imprecise. The bottom line is that the distance between the estimated market size and the minimum market size for success of your project is best measured in orders of magnitude, not percent. If it’s not that large, you’re threading a needle.
In other words, it shouldn’t really matter if the number of bulldozers sold in a year is 1.7 million or 2.1 million – it’s over a million and less than 5-10 million. If the project works at 2.0 million and fails at 1.6 million, do you really have a viable project?
What to do when the boss expects more
Folks who are hunting down market size information usually tell the same tale. Their leadership is ‘just sure it’s out there somewhere’ but no one can tell them where.
Folks, it may not be out there anywhere. If it was, and was easy to find, no one would hire anyone to go find it.
Measuring a market you have direct access to is hard. Very hard. Markets are dynamic things, and measuring takes time. It’s like measuring the weight of a newborn baby using a scale that takes two weeks to take a reading.
Even estimating a market is expensive and time consuming, and very few people do it for free outside of the government. That someone needs it doesn’t mean someone has gathered it, or even that it can be gathered in the way you expect.
The next time you’re confronted with leadership that has unreasonable expectations, try this experiment. With two-four of them in a meeting, ask innocently about some stat – last months’ sales, the cost of a product, the average income of a region, whatever. Just make sure it’s not something with clearly published numbers.
You will get as many answers as experts, and then you’ll hear the discussion about the various answers. Bill’s number doesn’t include direct labor…Sally’s figure does the currency exchange differently…Frank didn’t include shipping. Whatever the differences, the point is the same: If we can’t accurately measure and define a number we have complete control over, how can we expect to accurately measure and define quantities we don’t have control of?
Remember that the goal is to make a decision to do something. Data provides assurance not insurance.