In part 1 of this 2 part post, I explained how your CRM can be useful for buyer research, and how to find a list of qualified buyers from the thousands of people in your CRM. In part 2, I go deeper into how to analyze your list of buyers and build rough personas based on this information.
While each data set is unique to your company, sales process, CRM configuration, and conventions, there are typically seven pieces of information that we try to extract from a data set:
1) Buyer Titles
Probably the most obvious field for analysis in your list of buyers is their title at the organization. As similar titles generally have similar responsibilities, problems, and goals, this is a good place to start in making generalizations about the people who buy your product.
Analysis of this field isn’t as straightforward as it might sound, as any given title may have hundreds of permutations. For example, the title, ‘Director of Market Research’ is functionally equivalent to each of the following:
- Director, Market Research
- Sr. Director of Mkt Research
- Market research Director, North America
- Dir. of Customer Insights
…and so on. This makes a simple pivot table, which may be comprised of 75% unique entries, unrewarding. For most of these titles, you’ll never come across exactly the same one twice.
Instead, you’ll have to parse the field for seniority level and functional designation to get more digestible buckets for your titles, using a FIND function in excel.
For instance, searching for the keywords “director” and “dir.”, we can find the proportion of buyers in your data set that are approximately directors. Likewise, searches for “market research,” “mkt research,” and “customer insight” will tell you the proportion of buyers in the market research department of a company. Adding additional keywords should allow you to build a list or pie-chart of the seniority and functional category of buyers in your database, and give you a strong hint of who to pursue later on in primary research.
In this step, it’s important to find a balance of granularity and digestibility. The more granular you get with the seniority or functionality, the more difficult it will be to draw generalizations from this information, which, in the end, is the entire purpose of this exercise. But go too broad and you may dilute your conclusions.
You may want to cross-tabulate this information with company size, whenever available. This, for instance, might give you two separate sets of pie charts for SMB and Enterprise, where (more than likely) your titles skew higher in SMB.
2) Generalizing Responsibilities & Backgrounds
Once you’ve determined the most typical titles of your buyers, we like to choose a subset of those titles for deeper analysis in LinkedIn. Find their profile, and record some combination of the following pertinent information:
- Current responsibilities
- Previous job
- Previous responsibilities
- Years in current role
- Years at current company
- Years in the industry
- Age (22 + years since they graduated college)
- Major in college
Once you’ve collected the information, come up with the 3 or so most common answers to each of the questions (or averages, for ‘years’ fields).
This is a time-consuming and very qualitative process, so you’ll probably only want to do it with a small subset of your overall data set. However, reviewing these profiles can be extremely instructive into what your buyer typically does, the vocabulary they use, and the path they took to where they are.
3) Calculating Win Rate
While customers are an important piece of your data set, it’s important that you include lost sales opportunities as well. That’s because identifying and avoiding people who look like buyers but aren’t can be extremely valuable to your go-to-market organization.
If you’ve built your data set on contacts and matched that to the account status (i.e. Customer vs. Lost Prospect vs. Cancelled Customer), you should be able to calculate an overall win rate, and then cross-tab this based on other factors (like your seniority designation from #1) to understand where your most and least-likely buyers will come from.
For instance, this might tell you that your win-rate on Senior Directors is 50% higher than your win rate on Directors, and that a market research buyer is twice as likely to result in a win as a person from the finance department. This is an extremely important information when looking for prospects for an enterprise sale.
4) Calculating Incidence
Win rate is important, but it’s only half the battle. You may come across a title that has a very high win rate… but only shows up in 1% of deals. Is this a good title to go after, or are you chasing a ghost?
For this reason, we like to calculate incidence as well, to be displayed in tandem with win rate. Titles with very low win rates and very low incidence are pretty irrelevant, and titles with high incidences and win rates are obviously good targets. But titles with low incidence but high win rates could indicate one of two things: either they’re a really good title that you haven’t been looking for, or it’s a title that doesn’t typically exist in the segment under consideration. You may want to dive deeper into these titles by speaking to your sales and marketing colleagues to see which scenario is the case.
5) Segmenting by Product or Use Case
If your CRM includes a picklist for the type of product(s) that your buyer is considering, this step is a pretty simple matter of segmenting your data set by product and looking for patterns in the titles that buy those products.
If you don’t have this information, you may be able to approximate this by parsing the Opportunity Name. Many of our companies use a naming convention of “Account – Type Tier (Upgrade),” so an opportunity called “Walmart – Enterprise 1Tb” is Walmart’s first contract for the enterprise product, with 1 terabyte of storage.
Using a find function like we used in Step 1 to parse the title field, we can segment our data into storage and product tiers, to test whether a different type of buyer purchases each one. If the titles for each product are similar, then you don’t need to pursue this angle, but if the titles are dramatically different, you should consider using this as part of your buyer segmentation criteria.
6) Determining Objections and Competitors
Qualitative information about why buyers make the decisions they make is probably the most impactful thing you can learn from buyer research, but it’s also the hardest to find in your CRM. Still, it’s high value information that is worth spending some effort to find.
The first place to look is an “objections” or “competitors” field that some of our portfolio companies use to keep track of this information in a deal. This can be used rank the most common objections and competitors, optionally cross-tabulated with other characteristics such as title or product.
If your CRM doesn’t have these fields or they aren’t sufficiently populated, you may be able to constitute these fields yourself based on call notes. Like Step 2, this is a time-consuming process, so you’ll likely only want to do this for what you determine is a fairly typical subset of your data.
One CRM best practice that many of our portfolio companies follow is to scan signed contracts into their CRM for archiving purposes. While these contracts don’t usually contain much additional information versus what is already included in the Opportunity details, they do often include one important detail that’s often not readily available: who signed the contract?
Extracting this information is a manual process and shouldn’t be pursued for every closed deal. If, in your early results, the signature is the same as the person you recognized as the buyer, you can probably stop pursuing this angle rather quickly.
In our experience, however, the signature often belongs to the buyer’s boss. If this is the case, you’ll want to investigate how involved they were in the sale when you conduct primary research.
Bringing It All Together
Ideally, your CRM analysis will give you rough information to pick out a few distinct personas to focus on. Remember, the ultimate reason we’re creating multiple buyer personas is so that we can treat them differently during the sales and marketing process; so combine personas if you have no reason to treat them differently.
For each persona, this analysis should give you some or all of the following information:
- A list of typical titles
- The segment(s) this buyer applies to
- Their responsibilities
- Their typical background
- How to prioritize them relative to your other buyers (based on occurrence & win rate)
- The product(s) they’re most likely to buy
- Their most common objections
- The competitors they’re likely to consider
- Who’s approval they need to become a customer
Using this rough persona as a starting point, you’ll be much better equipped to execute the meat of a buyer insight project: conversations with actual buyers.
Just as important, this exercise will let you know what information is missing or unclear, so that you can focus on these areas more during primary research.