Big Data, Small Business

Last week, we talked about questions.

Questions tend to produce answers and more questions, which can result in a pile of stuff that overwhelms a small business.

As a business and client base scales, these questions produce data that you can use for guidance, decision making and to ask even better questions. Again, this can result in a pile of stuff (data, in this case) that overwhelms a small business.

A common reaction to this phenomena is to ignore the data, or to be so overwhelmed by its volume that you can’t discern anything from it. Entrepreneurs tend to want to do it all and if they can’t do that, doing part of it seems like a failure. It isn’t.

Identifying your big data

Let’s look at one of the questions from last week’s post and see which ones are likely to produce decision-making data.

How does this impact our key performance indicators? Examples: cost per lead / new client / sale / deployment, support load, lead time, etc.

This implies that you already know your cost per lead, cost to acquire a new client, cost per order/sale, cost per deployment, average lead time per product/service and the support/customer service load your products and services require. Not gut feel, but actual numbers.

Actual numbers are important because our gut is often right when it comes to strategic decisions and the like, but it seldom has a clue when it come to numbers like cost per lead – particularly if you’ve never watched it.

Lead cost, sources, media and campaigns

For example, what impacts cost of a lead at your business?

Lead source is a good place to look.

You might get leads from referrals (cheap and strong, warm leads), from local TV ads, from local newspaper ads, from different media in your education-based marketing, from the phone book (yes, some businesses still depend on those leads), from direct mail (likewise, still quite productive if used properly), from your website, mobile app, and so on.

Each of these have different creation and distribution costs. Each will produce a different lead flow, much less volume and types of client. While in the beginning, you’re likely to lump all of this data together, at some point you need to break them out by media and eventually, by campaign.

You’ll want to do that so that you can answer questions like this:

  • How do you know which media produces the most profitable clients?
  • How do you know which campaign (and on which media) produces what number and type/quality of client?
  • How do you know if a particular campaign works well on one media, but terribly on another?
  • How do you know which media (or campaign) tends to produce clients that are high maintenance to the point that you tend to fire them or not accept them in the first place?
  • How do you know which media produces the best (however you define that) clients you have? Is there a specific type of campaign that does this?

From time to time, an owner will tell me that their businesses doesn’t do any marketing so this kind of thing doesn’t help their business. If that’s really true, you’ll usually have referral sources that produce more and better leads than referrers do.

Would it be helpful to know who is sending you the best referrals?You probably have a gut feel on this, but are you sure that it’s accurate?

Thinking back on those questions

Given the detail on the one question of cost per lead, you can see how this can become overwhelming in a hurry. Don’t fall victim to that. Take it a step at a time.

You may start with another metric. Cost per lead is important for almost everyone, but it isn’t always the best place to start.

When you ask questions like “How did the pilot program go?” – it might provoke follow up questions about the data collected during that pilot which would support the “How did it go?” question.

If those answers aren’t backed with data, then that might provoke you to add data collection to your pilot projects in the future. This will take more time but it will produce better answers that don’t depend on gut feel or a need to be right.

Better answers are what we’re looking for.