Start with very simple analytics

For years I’ve been setting up daily metrics spreadsheets for companies that I own or consult for. I have learned how critical it is to 1) set goals, 2) track results every day, 3) look for new channels continually, 4) and test new creative regularly.

When a startup company looks at a mature internet company’s analytics and daily spreadsheets, it can be overwhelming.

I want to advise all internet startups to start simple. Track spending and new customers. Don’t get overwhelmed at first with all the complex details of a mature analytics program. Start at a high level and drill down later.

But first, just track simple things like daily spending, new customers, and cost to acquire a new customer.

If you focus on acquiring new customers through internet marketing and make sure that the campaigns you are running are acquiring new customers at a reasonable level, then you can increase the volume and optimize the campaigns with time.

But if you don’t even know your cost to acquire a new customer there is no way you can make good decisions about internet marketing.

I think this is the first key metric that internet marketers should track.

I think it is the key metric that good direct marketers have been focused on for almost 100 years.

In the late 90s it seems that internet entrepreneurs had a sort of arrogance. We classified people as getting it or not getting it. All you had to say about someone is “he doesn’t get it” and that communicated volumes.

We thought we were the first advertisers/marketers in history who understood how to track the results of our spending and make sure that all of our advertising was actually profitable. We knew that web analytics made this possible, whether we used home grown systems or third party systems. Unlike most advertisers who knew that half of their advertising worked, they just didn’t know which half, we could track our spending on a daily basis at the campaign level. We knew what worked and what didn’t. Or at least we thought we did. (Even when spending millions a year on online advertising and analytics, we were always plagued at MyFamily.com with a very large “other” bucket.)

When I discovered the book “Scientific Advertising” by Charles Hopkins, published in the 1920s, and learned that he had pioneered couponing and all kinds of direct response tracking, I was humbled by how brilliant some of the pre-internet marketers really were. Even now, direct mail and direct response gurus exist who generate millions of dollars of trackable spending. They really are scientific about their work. Unfortunately, there are still many “brand marketers” who are creative but really don’t know how to generate revenue. They hide behind the fact that much of their advertising spending can’t directly connect to revenue.

I remember how incredibly disappointing the $10 million MyFamily.com TV advertising campaign was back in 1999-2000. Deutsch did the creative and millions were spent with almost no impact. But in a pre-IPO situation that we were in, venture backed companies were encouraged to spend money like crazy to build a brand. (I personally preferred Mary Meeker’s advice on branding–forget branding and spend money to acquire new customers–that is really how to build a brand.)

As I introduce analytics to my students and Academy members, I worry about overwhelming them. After all, Omniture can generate hundreds of thousands of reports.

So I’ve been thinking back to my first months in online marketing and the simplest and most important daily reports that I required.

It comes from asking two simple questions: “How much did we spend today online” and “how many new customers did we get.”

With those two numbers, you can calculate the cost per acquisition number. Track that over time. See if it is going up or down. Set a target and have a flexible budget that says as long as we are acquiring customers for this target amount or less, we can keep spending more money.

Once you have this simple number and track it over time, then the web analytics can help you divide all your marketing efforts into different channels, which you can start tracking separately. Again, you should do it on a daily basis.

So your overall numbers might be:

Costs New Customers Cost Per Customer
11/5 $250 10 $25.00
11/6 $275 15 $18.33
11/7 $300 12 $25.00

Then, you want to start tracking individual marketing channels (Google vs Yahoo vs MSN, each email campaign that you run) to see what the CPA is for each one.

If your CPA is good, then run more and more campaigns. If your CPA is not good enough, test new ads and landing pages until you can get a good enough CPA.

The holy grail for internet marketers is to have a low enough CPA and a high enough gross margin that each new customer you acquire actually generates profits to the bottom line.

If that happens, then follow what I call the “blank check approach to internet marketing.” Spend as much money as you can on any marketing program that brings you new customers that generate a profit on the first order.

If you lose a little money on your initial order (most ecommerce companies do), then you have to hope for repeat orders in order to generate a long-term profit on your marketing campaigns. You can start developing a “life-time value” model that tries to predict how much you can spend to get a customer, knowing that over time the repeat orders will help you turn a profit.

I’m hoping to set up a public Google shared spreadsheet that internet marketers can use as a model for their own daily CPA tracking.

I have several, but I’m afraid they are too complex. I’m looking for a simple one that I can share as a template.

If you know of one and are willing to share, please let me know.

71 total views, 3 views today

4 Comments

  1. Kevin Schiess

    Paul,

    Thank you for the great blog! A shared spreadsheet would be SO helpful for me. I look forward to seeing it and thanks again for all of your insights.

    I heard David Antley speak today at the Rodizio Grill about internet marketing. He is a very intelligent marketing guru who got me very excited about online marketing and its potential for our little company.

    I asked him about dynamic keyword insertion that I read about on your blog. I implemented this in my ad campaign. He said that the average search is much longer now than it used to be and therefore you would most likely be cutting off some of your words and he didn’t recommend it. What are your thoughts on that?

    Also, he recommended a book called The Search which is a good book for a novice like myself. Have you read this book?

    He also mentioned multivariate testing and said that this will become the standard in online marketing in the near future. Do you agree?

    Kevin Schiess

  2. Kurt Haug

    THIS is where you really shine, Paul!

    I love the good, solid, simple yet POWERFUL advice you give folks based on personal experience as well as independent research.

    Those of us who lived through “the bubble” owe it to those we work with to encourage independent thinking– and to NOT just “drink the Kool-Aid”…

  3. Adriano

    I would like to see an example of the spreadsheet you mention since I would like to start tracking these areas. Thanks for the great information you provide, it’s very helpful.

Leave a Reply

Your email address will not be published. Required fields are marked *