Yesterday, we discussed a relatively simple methodology to identify the predominant industries of one’s customer base and also market sizing for same. We called it: Determining Industries and Market Size – The Cheap and Cheerful Method.
Now we move onto a somewhat more elaborate analytical effort (and it will cost a bit of money) but it will give you substantially more detailed information to base your planning upon. We’ll call this one:
Determining Industries, Market Size, Firmographics and “Best Customer” Profile
Let’s dive in:
You want to generate a list of ALL your very good present customers. Eliminate dormant customers, poor payers, perennial money losers (ones that you keep just for the cash flow but would otherwise “fire”) and any other customers that, for whatever reason, you could live without. Your list should now represent the customers that you would consider to be yourbest customers. Put them all in a spreadsheet. For each customer, include company name, full address with zip code, phone number and you might as well include any other data at hand like their sales rep’s name, sales, average sale (if applicable), source (how they came to you), etc., all in separate columns. Presumably, you would LOVE to have hundreds more customers just like them.
2. Indentify the industry to which your customers belong.
This next step is the part that will cost some money because you will need to engage a “data cleansing and append” company. These companies maintain massive databases of virtually all companies in the U.S. (and internationally) with rich information on each company. They provide data cleansing and enrichment services and will also sell lists of companies for marketing purposes.
They will take your database and, in general, do the following:
Apply an address correction software to your data (this helps with matching)
Cost: I researched one provider of this kind of service, a company called Melissa Data. They have a minimum order fee of $500 but they charge based on the size of the initial database you give them, the elements you want to append and the percentage match rate they ultimately achieve.
Let’s show an example using published pricing from Melissa data: You submit 10,000 records and you want employee size, sales volume and NAICS code appended to as many records as possible. They achieve a match rate of 40%. Their cost would look something like this:
|For 10,000 records||Match rate||Appended||Price per 1000||Cost*|
|Sales volume range||40%||4,000||$60||$240|
* Example: 10,000 x 40% x $60/M = $240
You could call this a pay-for-performance pricing model because they only charge when they get a match (in this case 4,000) not on the entire input file (10,000).
Match rate: You may wonder why only 40% or so are matched and this may seem low to you. However, in the world of basic data matching, 40% is not a bad result. It is possible to achieve a much higher rate, perhaps as high as 75% but that would entail a much more elaborate matching process that comes with a price tag that is many multiples of the solution I’ve outlined above. Also, in this example, we matched 4,000 and having that as our sample set should deliver a statistically accurate outcome. So it’s not absolutely necessary to have a high match rate to get your desired outcome, which is a profile of your best customers.
Confidentiality: You may have reservations about data security and the confidentiality of your customer information in the hands of a third party. What I can tell you (and I worked for Dun & Bradstreet for 10 years) is that companies who do this kind of sensitive data work go to great lengths to handle a customer’s proprietary data with utmost care and security. Frankly, they have to because if there was even a whiff of data security issues, it could do serious harm to their business. In short, they have maximum incentive to handle your data with extreme caution.
In this example, you would get your full database back and 4,000 of them will have this key firmographic info appended to it. This is your raw database to start the next phase of the work that we will discuss …tomorrow! That’s right folks, please tune in tomorrow to read the final article in this series of three. OR …
To see the entire text of these three articles RIGHT NOW, click on the following link: