Understanding Metrics: Lead Generation

Dec06
By Joe Izenman

Measurement Strategies for Websites that Focus on Sales Lead Generation

This is part four of our series on Understanding Metrics. For a high-level overview of the topic, check out the introduction blog.

The Goal

We’ve talked a bit about the joys of measuring success by normal transactional retail, but what if your business is a little more hands-on? For larger ticket items and contracted services, a sale is the product not of “see thing, buy thing,” instead it’s a longer back and forth of discovery, proposal, and negotiation. As a primary point of first contact, your website plays a significant, complex role in this interaction.

The Metrics

Let’s start in the same place as the last few goals: revenue. If you’re making more money from leads that your website has generated, then that’s success, right?

To keep this measure meaningful, we need that direct link from website to sale. On the surface, this is straightforward. If the client reached out via your contact form, then you have a connection. There are, however, two subtle distinctions:

  • Was the website the driving force, or simply the contact channel? If someone hears about you by word-of-mouth and comes to the site looking for more information, you may want to reduce, or lessen, the calculated impact of the website.
  • Conversely, interaction with the website may be a key factor for a customer who ultimately reaches out by phone or email.

Both cases necessitate surveying leads as they come in. This can be as simple as asking “Where did you hear about us?” on the form or during a conversation in the discovery process.

This paradigm also shares elements with in-store offline revenue: There’s a lot unaccounted for between the initial contact and final sale such as varied salesperson quality, dramatic change in client needs, etc. Because of these variations, it’s important to develop a success metric for your site that is somewhat divorced from the outcome, capturing the influence of the website at the point where the process leaves its purview. There are a couple of options for this kind of insight:

1. Lead Volume. Naively, total leads generated is a positive indicator. The more people who are interested in your services, the better. This is vulnerable to a couple of issues:

  • Spam: There may be a lot of garbage coming in via an open web form. A strict count without filtering can severely skew your numbers.
  • Capacity: Aiming for a high volume typically assumes a low revenue-per-lead. You’re either hoping to make a lot of low-level sales or find the one diamond. Both strategies need a lot of hours to sift through the volume.

2. Lead Quality. Instead, we may aim for high lead quality, rather than volume. The specifics vary by organization, but, by analyzing all of our data on past leads, we can construct a model of sales likelihood. Assuming an average salesperson, and some degree of uncertainty, does this lead have the characteristics of one that should make a sale? Put another way, if this same lead came through a hundred times, what average amount of revenue would be expected?

A model like this is constantly evolving. While its goal is to be predictive and judge a lead before the sale is closed, it can also be updated post-sale with the final outcome to improve the next prediction.

Next Up

As much as revenue might be the end business goal, there are plenty of websites out there that aren’t intended as a direct feed into sales. Next week, we’ll start looking at the uses of general user engagement as a positive force for your business.


Data Science

Back To Feed