Do you really know who your customers are?
The key to scaling a B2B business is having a highly repeatable sell. This includes 3 components:
- A small set of well defined problems and corresponding solutions
- Well defined customer profiles for the companies, industries, decision makers, and users
- A well defined sales and marketing pipeline process and engagement model
Let’s talk about #2 today. Do you really know who your customers are?
One of the most commonly used B2B lead generation strategy is to find more prospects that look just like your existing customers. Ask your sales and marketing team what does your customer look like in terms of:
- Company size and maturity
- Who are the first points of contact and discovery
- Who are the decision makers, budget owners, and influencers
- Who are the day-to-day users
You are one well oiled sales and marketing machine if you have well documented answers for all of the questions above. How did you develop these profiles? Was it based on interviewing the sales team or was it based on analysis of your customer data? Chances are, your customer profiling analysis was developed using a combination of hypothesis and field intelligence, and very little hard marketing data to back that up.
For example, are your customers CIOs, directors, managers, or analysts? Are they in IT, support, marketing, or in the line of business? Hypothesis are often colored by our perception of ourselves, that can lead to customer profiles that are more who we think we should sell to than who actually bought. For example, you may think your users are from engineering, but in reality they are from support. Field intelligence can also be misleading and is often based on anecdotal evidence. Every B2B sales rep knows you are supposed to “call high”. Really? Is the CIO really the person that will make the technical decision on say a dev-op tool?
The only way to really accurately profile your customers is through marketing and sales data analysis. To do that, you need to have fairly clean customer data and the right customer profiling tools to breakdown the data. Data clean up is a whole big topic for another day.
When it comes to analyzing customer data, the minimum you should do is find the most common values for job title, job function, seniority level, company size, industry, SIC code, and possibly others. You can start with an analysis like the Rank Report in Openprise. Here is an example of ranking the number of contacts with specific job titles:
This is insightful. For example, you may not know that DBA is in the top 3 of your most common user titles. However, there is a fundamental problem with this first analysis: there are too many job title variations in your data, not even accounting for spelling variations and typos. In the above example, “Manager, Information Security” and “Manager, IT Security” are counted as two different titles. To further illustrate this problem, let’s take a look at the job title breakdown of the entire database using Openprise’s pie chart analytics:
These top 20 job titles make up 6.6% of your customers. That big 93.4% is “others”.
You can overcome this problem by normalizing the job title data. That is a hairy topic for another blog.
One simple solution to this problem is to do a word count analysis on the job title. Here is an example using Openprise’s word count analytics:
By counting the frequency of appearance of individual words in people’s job title, we are getting a much more clear picture of who our customers are. They are first and second level managers and they are mostly in the information security function. There are relatively few people from development, engineering, and operations.
You can do similar sales data analysis with industry and job function data.
You can refine this analysis by focusing on first contacts, support contacts, decision makers. You now know who finds your company, who influences the buying decision, who writes the check, and who uses your product.
When it comes to customer profiling analysis, hypothesis and field intelligence is a good start, but back it up with data and you may be surprised to what you find.