A Marketer's Odyssey

There comes time in most marketer's lives when the excitement of converting customers into leads begins to wear thin and he's forced to confront the sales numbers in front of him.  

The leads, you discover, though cheaply they may have came, are now falling out of the funnel at an alarming rate.  

Sales weren’t being made.  Money is flying out the door.  It's time to sit down and have a serious conversation about lead quality.

These conversations happen differently in different organizations, but usually they involve the whole marketing and sometimes the sales team to talk about what they see in leads coming in.  

I ended up in a conversation like this one last year.  The question on the table was this:

How can we find better quality leads in our paid social channels like Facebook and Instagram?

In this case, we were selling a very expensive and sexy product with a very long sales cycle.  This meant that a lot of people came into the top of the sales funnel, and a much smaller number came out the other side and turned into revenue.  

Once they realized how expensive the product was, the excitement wore off the they fell out of the funnel.  There was lots more we could do to better nurture these leads, but my goal was to increase the percentage of people coming in that could actually afford our product.  

A Siren on the Rocky Shore

I decided that the best place to start looking for an answer was to get very serious about the audiences we were advertising to on Facebook.  A few tests were up and running when I my eye lit on something stunning.

Facebook's income targeting menu

This was the answer to my problems.  Income was a huge barrier to entry for our product, and one that we’d had trouble replicating with proxies like “College Educated” or “Likes both pages about wine and yachts”.  It was time for a test.

The Test

The lead form on our website collected all leads self-reported income ranges, and I tracked the leads through to our CRM with UTM tags so I could associate leads with particular ad groups.

A few weeks later, I went in to look over the results and my jaw hit the floor.

All leads from the test using Facebook's reportedly 125k+ income targeting

55 of the 90 leads had an income of under 40k

35 of the 90 leads made over 40k in the 125k income targeted group

WTF?

I’d tested an ad group with no income targeting against an ad group using Facebook’s income targeting to isolate people who made over 125k a year in the United States.  

This all made sense- until I find out that of the 90 leads in the 125k+ group, a full 55 of them reported an income below 40k.

The income targeting on the 125k+ income group

That meant that only 39% of our leads from the 125k+ group made over 40k in a year- the exact same percentage of people who made over 40k as in the untargeted control group.

On top of that, the CPC (cost per click) were also more expensive for the 125k+ group.  That is to be expected if they’re doing what they’re supposed to be doing and bringing in better leads.  But they weren’t, and Facebook was charging us a good amount more for them.

What does this mean?

For businesses selling high price-point products, it means the same thing it’s always meant- you’ll have to keep using proxies to help you guess an audience’s disposable income.  Lookalikes, reductive targeting, proxy fan pages- there’s no end of good places to start.  

It also means that Facebook’s markers for judging a person’s relative wealth is astonishingly poor.  It just goes to show that sometimes it’s easier to judge a book by it’s cover than by running thousands of “personality inputs” through an approximating algorithm.