Predict the number of future clients – howto

Predict the number of future clients – howto

This post is also available in: English Français (French)

Total addressable market analysis – what does it mean for your plans

To predict the number of future customers we use Total Addressable Market analysis, TAM. The technical term TAM refers to the number of possible customers and the number of products that they will buy as a whole. This is the full market. If you sell cars, everyone can be your customer. There are competitors, and not everyone is going to buy cars from a new producer. So the number of cars you are going to sell is smaller like the market for automobiles. If you want to sell Yoga classes, everyone is a potential customer as well. The vast majority of clients for Yoga classes are women, which decreases the number of clients. People who need to go over 30 kilometers for the workout are unlikely to buy your lessons. For a year now, Covid-19 has been a big problem – either lessons together in a gym are prohibited, or older clients do not want to come due to fear of infection.

The total addressable market narrows down to the servicable accessible market.

These are the customers who may be ever hear from you and suscept you as your potential supplier. You have to be able to sell your products to them. This is much closer to the reality and is one step calculating your future sales.

The third and most important technical term is the “_S_ervicable _O_btainable _M_Market”, SOM. These are the real customers, the ones you can serve and the ones who want to buy your products. They need to know your products and services.

The math

For calculationg, demographic statistics are important. These data indicate the number of people in the target region.

How to find the number of potential customers? The primary source is Wikipedia (really) and the governmental Statistics Office. Their data is normally free. The state administration places great value on correct data. They use this to distribute taxpayer money.

You can find more than population data there. The number of dairy cows is also there, and a lot of the data by economy. So it is possible to predict the number of future clients.

Boutique market researchers sell premanufactured reports

Using google to search for the data you arrive at several websites that are trying to sell industry reports. The normal price for these reports is between 800 and 3700 Euros. On my Fiverr account, a website that acts as an intermediary between freelancers, a lot of requests are aimed at selling these reports for less, say 80 €. The reason the prices are so high is that the authors use the most expensive data, for example at Bloomberg.

Is it necessary to pay more? For small entrepreneurs who offer specialized products, the answer is no. The reason is: it doesn’t say anything which we can use to predict the number of future customers.

Data Quality needed to predict the number of future clients

To estimate the SOM, industry associations offer a lot of data. Perhaps this data is old, but there are no or little fees.

The internet offeres a lot ot data about what is popular with users and what they think an want. That data is hard to find and to be familiar with with web evaluation and web scraping ist helpful. In 2020 I worked on a market study for flower pots. There was a lot of data coming from the associations of the friends of the gardens. They show me that all people love their garden and their flowers more and more. Are the data true and dear to the seller of the flower pots? How much money do they want to spend on the jars? What appearance will they like? Lots of work to do.

Quick and only a little dirty

For very quick results it is possible to do an analysis of reviews on Amazon and other sellers. There is software to do this. Youtube is full of videos that show you this technique. Here are some basics how to analyze that (German) and how I do it.

Published by Johannes Winterhalter
Gaining numbers from multiple sources and transforming that into meaningful and comprehensive results - I do that since DBase III in 1988 and am now into data lakes, SQL, R and survey management.