Web metrics: The good, the bad and the useful
I may be wrong. I frequently am. But there are three Web metrics people seem overly concerned with that I just don't worry about.
The first is open rates. Since both my online magazine and e-mail marketing messages are text, I can't measure open rates.
I could convert my text e-mails to text in an HTML shell, which would enable me to track open rates. But why? As long as e-mail is profitable, generating a lot of sales, what do I care how many people opened it? After all, in direct mail we only know how many people responded by returning the order form with payment. We have no idea how many people opened the envelope or whether they read the contents.
The second metric I don't care about is page views: what pages of my Web site are visited most, how many minutes the average visitor spent on each page, which pages people returned to and so on.
Again, my concern is whether the Web page can convince visitors to order the product. If they don't order, it should at least get them to give me their e-mail address and opt into my list.
The third thing I don't care about is complaints. A reader recently asked me, "How many complaints do you get from AOL users?" Why would I track AOL users separately from the rest of my list?
Look, I don't like complaints. If someone doesn't like an e-mail I sent, I take it seriously and respond thoughtfully and politely. But it doesn't keep me up at night. Just as they can change the channel if they don't like what they hear on the radio, they can opt out of my list if they don't like my content - and they should.
There are five Web metrics that I watch like a hawk. More than that, I live and die by them.
Click-through rates. When I send an e-mail to my subscriber list, how many of the people on the list click on the URL link to the landing page? Total clicks, unique visits to the landing page and the click-through rate, which is a percentage, get measured. For example, if 800 people on my list of 40,000 click, that's a 2 percent click-through rate.
With e-mail marketing messages sent to your house list, click-through rates can range from 1 percent or 2 percent on the low end, to 10 percent or 15 percent on the high end. These click-through rates are for e-mails selling a product, not e-mails inviting the reader to get a free white paper or other free offer.
Conversion rates. Conversion means converting people who visit your landing page into buyers. If you generate 1,000 clicks to your landing page, and 100 of those people place an order, your conversion rate is 10 percent.
Conversion rates can range from 1 percent or 2 percent on the low end, to 10 percent or more on the high end. Inexpensive products generally have higher conversion rates, while an e-mail promoting a big-ticket item can be profitable even with a conversion rate below 1 percent.
Combined, the click-through and conversion rates determine how many units you sell. For instance, if you get a 3 percent click-through rate on a list of 40,000, you get 1,200 clicks to your landing page. If your conversion rate is 5 percent, the e-mail generates 60 orders.
Gross sales. This is the number of orders generated by the e-mail multiplied by the selling price of the product. On the above example, 60 orders for a $29 e-book generate gross revenue of only $1,740. But for a $249 product, 60 orders produce total sales of $14,940. In my little Internet business, the former would send me into a deep funk, while the latter would have me popping a Champagne cork.
Opt-out rates. Every time you e-mail to your list, a certain small number of subscribers decide to opt-out or "unsubscribe" from your e-mail list. That's a bad thing, because unless you do something to generate new subscribers, your list will gradually dwindle to nothing.
What is an acceptable opt-out rate? My average e-mail to my list of 40,000 causes about 20 people to unsubscribe, which translates into an opt-out rate of 0.05 percent - half of one-tenth of a percent. I think you can live with an opt-out rate of 0.1 percent per e-mail, but more than that and your list will shrink too rapidly.
Dollars per name. Dollars per name is the dollar value of each name on your list - the amount of revenue per name. If you make $200,000 a year in online sales from your subscriber list, and you have 40,000 subscribers, your dollar value is $5 per name per year.
Why is this important? Because the various methods of building your online list - pay-per-click advertising, co-registration, banner advertising - all cost money. Your dollars per name tells you how much money you can afford to spend to acquire new names for your online list.
For example, if your dollars per name is $5, you certainly can afford to pay $1 to add new names to your online list: on average, you'll earn five times your investment within one year.
Say you use pay-per-click advertising to drive people to a landing page where you offer a free report as an enticement to opt into your online list, and half of the people who visit the page accept the free offer and subscribe. You can therefore afford to bid up to 50 cents a click, since it takes two clicks, costing a dollar, to get one sign-up.
Now, general advertisers, brand marketers and business-to-business marketers generating leads rather than direct sales may care about many other metrics. But for the direct marketer selling a product online, the five metrics listed above are the most important metrics you can track.