Return to: GROK Dot Com 12/01/2001

Déjà vu: Engineers and Customer Metrics

It happens every time. We at Future Now start talking to a client who's keen to improve conversion rates, and we ask if there are any site-tracking statistics we can evaluate. "Oh, geez, yeah!" the client exclaims. "We've got tons of tracking data." And they do. Reams of it. Except none of it is particularly helpful to us because none of it really reveals what is actually happening as customers use the website. We always find ourselves back at Square One - it's the inevitable result when systems are assigned higher priority than customers.

My friend Jim Novo, former VP of Programming & Marketing at Home Shopping Network, has seen it all before, repeatedly, and is suffering a serious case of déjà vu. I'd like to share his thoughts and experiences with you.

Dear Grok,

If you build it, they will come. They just don't stay. This is the legacy of "engineering-first" innovation in consumer services, where every metric is devoted to tracking how the physical business is running, and none to tracking customers. How hard would it be to set up customer reports in the beginning? But it never happens. What always happens is there is a spurt of intense growth (accompanied by hallucinogenic stock market valuations), then the business flatlines and rolls over on its back. I saw it in cable TV. It Happened again in TV Shopping. Now it's happening with the Internet.

When I joined the cable industry in the early 80s, it was all engineering-driven. The "truck-chasing" days, we called it. All you had to do was show up with a cable truck, start stringing cable, and people would run out of the house yelling, "It's cable! Cable is here! When can I sign up?" All sales-driven, and all about miles per day of cable, amplifier density, and number of "drops" and "traps". Everybody talked about drops and traps. Customers? Nah.

You know what a drop is? A line from the main cable plant to a customer household. I could get reams of reports on drops. Nothing on customers.

Me: How many customers do we have?

Engineer: We have 12,239 drops.

Me: How many former customers do we have?

Engineer: Why would you count inactive drops? There's no hardware there.

Traps? These are the filters used to control who got HBO and who got Showtime. I could get reports on how many HBO traps we ordered and how many we had left. But how many had Showtime too? How many former HBO customers were there? No way to tell. All we had was the "net HBO traps" - I ordered this many, I have this many "in the barn", so the difference is how many HBO customers I have.

Me: Well, how many "dual" customers do we have, HBO and Showtime?

Engineer: We have 6,783 HBO traps and 3,872 ShowTime traps "in field".

Me: But how many customers have both?

Engineer: What difference does that make? Costs the same to install one or both, as long as we do it at the same time.

Aarrgghhh!

Metrics based on hardware, inventory and so on that mean something to the people keeping the architecture operational. So valuable to engineering. So useless to marketing. Who are my best customers? Who are my former customers? And they wonder why penetration had stalled at 40% and customer churn was growing every day. The good news was I fixed it. It took 6 years.

Then I went into TV Shopping. I'm thinking here's a business immune from the anti-customer reporting bias. An electronic business, built by IT engineers, completely and totally dependant on computers to operate. No computers, no TV shopping business. Perfect! Every nuance of every transaction captured. Time of day. Category of merchandise. Method of payment. A goldmine of data. Not like cable TV.

So I asked for reports. The first day, the stack of greenbar was over four feet high! I'm in heaven, it's a data-driven marketer's dream! I dig in.

Sales report. Sold 10,000 bracelets yesterday, broken by style, color, length. Awesome! How many customers bought the 10,000 bracelets. Hmm. Not on this report, must be on another one. Oh, look at this - sales per minute by category and by time of day. Wow! How cool is that? How many customers? What percent were new customers? Hmmm. Must be another report.

And so it went. Calls per minute, returns by price point, check fallout by time of order. Hundreds of metrics - every one of them based on a transaction: merchandise, capacity, throughput, inventory, logistics - and not a single one at the customer level! We're counting every neuron that's firing, but nobody knows what the brain is thinking.

That took 10 years to fix.

Now the Internet. I think, "Here's my chance. They'll listen this time, because, if nothing else, engineers are logical. I can show them the math." I write a book. Put up a website. Lay out the whole thing - here's exactly how you do it. Collect this data. Here are the metrics that matter.

And what do people decide to track? Hits, page views, browser types, files downloaded - machine stuff. As bad as cable TV. The net retailers tracked just enough to fulfill (barely) the orders - usually SKU's with no further detail. Worse than TV shopping! But customer level data? Nah. Furthest thing from their minds.

Why does this keep happening over and over? Is there any way to get in front of it? What will it take for the engineers to realize, in the midst of their heady glee with being new and innovative, that the same things always happen to these technology-driven services:

1. They don't set up customer-oriented reporting.

2. They don't know what customers are worth, or how they really use the service.

3. They way over-project long-term profitability.

4. They invest in infrastructure beyond what is needed.

5. The customer side catches up with the hardware side and the business rolls over nearly dead.

6. The engineers call marketing to fix it.

7. Marketing tells the engineers it's going to be a long hard road because they screwed it up so badly already.

Any ideas? Anyone? How can we stop them from hurting themselves again?

Sincerely,

Jim Novo

We're always thinking about the ideal tracking system at Future Now (see "The Ideal Tracking System,") and helping you rethink how you collect the data that will allow you to identify and remedy obstacles in your website's conversion system (see "Measuring Your Sales Success,"). We even offer a free, downloadable suite of calculators to get you started. Because at the end of the day, your goal is to build a business, not a house of cards. Right?

How do you prefer to be addressed:

Your email address



We Value Your Privacy!
Forward This Grok To A Friend!

Return to: GROK Dot Com 12/01/2001

© 2001 Future Now