Mental Checklist

Every time a visitor enters your site he/she runs through a mental checklist.  Some of the items on this list are:

1. Do I really need to buy this item?
2. Should I buy it now?
3. Is the item priced fairly?
4. Is this retailer credible?
5. Are the terms of purchase favorable?

Even a single “I’m not sure” to a question above could derail the sale.

As marketers we need to ensure:

1. These questions are answered satisfactorily.  Note: Just because the answers are acceptable to you doesn’t mean your potential customers think they’re acceptable.

2. These answers are easily accessible.  Note: Just because these questions are answered somewhere on your site doesn’t mean people are seeing it.  Use analytics to verify accessibility.

I believe these 5 questions are so important that every 2 months you need to draw up a list of reasons why people may answer “I’m not sure” and then enhance your messaging to fix it.

Whisper Marketing is an e-tailer specializing in exotic bird supplies.  Their site is popular but their visit volume represents a tiny sliver of US bird supplies market.  I am sure they would love introducing more customers to their site.

Here is a potential solution.  Who do exotic bird owners know?  That’s right, other exotic bird owners. should run a small scale experiment-

1. Pull up email addresses of their 100 most loyal customers (ranked by average order value, buying frequency and length of relationship).

2. Around early November send this list an email with a message that boils down to: because you are a valued customer we would like to send a box of 50 complimentary holiday cards.  Click here to pick your favorite design.

3. On click customers are taken to a page with 3 holiday card formats.  Card cover has a beautiful high quality picture of a bird (holiday themed) with a message on the inside.  Message choices vary from generic to funny.  Customers can even mix and match their selection of 50 cards.  The card itself has no branding except for the phrase in small font on the back.

4. By the 3rd week of November these customers receive a box of bird holiday cards.

5. They now send their cards to friends and family.

I’m not a printing expert.  I believe bulk buying 5,000 (assuming 100 shortlisted customers) + mailing the card box costs $7,000.  People who visit are redirected to  <—This long url tracks card marketing campaign in Google Analytics.  To customize for your site visit

200 days later the company logs into Google Analytics, accesses Traffic Sources –> Campaigns report and measures revenue, transaction, conversion and per visit value numbers.  The company already knows new converts have a lifetime value of 3 transactions.  Using this they calculate card campaign net profits.  If ROI is greater than 3x the program is expanded.

Related article: KuKuRuZa Gourmet Popcorn

Geo Mining

You are a gourmet chocolatier.  You open Google Analytics, access Map Overlay report and segment by Transactions.  On the screen appears a heat-map of purchase data.  States that are dark green have had heavy transactions and light green states have slow activity.  You can see Michigan is a strong base, Texas and California are fairly good but Utah and Indiana are poor.

Next you open your sales data and notice something interesting– when a new order happens in a city it often leads to more local customer orders.  This tells me word-of-mouth is spreading locally.

In the book Tipping Point author Malcolm Gladwell describes how phenomena grow slowly for a while and then explode.  The number of orders we’ve had in Michigan suggests we might have reached a tipping point there, but it’s clear we have ways to go in Utah.

Our chocolatier has a limited marketing budget and we want to get the most bang for the buck.  The company is willing to give a generous 50% discount for 1,000 new visitors.  We know Michigan has crossed inflection point so no need to give additional discounts here but I’d be willing to entice a visitor from Utah with a 50% discount.  Thanks for Geo IP we can target in this manner.

Have you heard of RipStiks?

For teenagers in my neighborhood this the rage.  I’ve probably seen 15 kids play around in them.  My hunch is this is a social phenomenon.  At some point four kids in the neighborhood got the board and then everyone had to have it.  The houses in my neighborhood fall within one zip code.  For the marketing manager at RipStik this is a gold mine.  She has access to the sales data and can sort that entire database by zip code.  Using this data along with available school district/income/family demographic data she can identify zip codes where they’ve sold less than three units.  That list of zip codes is where we need to offer visitors discounts using Geo IP.  Once we reach four orders from a zip code discount is turned off for that zip code.  The idea is to only offer the discount till we reach inflection point.

Related article: Illinois Residents Pay No Sales Tax

Did You Find What You Were Looking For?

With on-site search etailers can measure result page effectiveness by observing Search Exits and Search Refinements.  However, the big limitation with analytics is that it only answers the “what is happening?” question.  To get to “why is this happening?” we need to make a small code tweak and display this below result page (borrowed from does this too…

When Data Lies

Just because your metrics don’t show the customer has abandoned doesn’t mean she doesn’t hate your guts.  I’m not terribly fond of LegalZoom and have a fleeting feeling their agents are pushing only their most profitable products.  That said, I’m unwilling to shop around for a number of reasons-

1. I already have an account and transaction history with the company.
2. How can I be sure the next company would be any different?

At data level I can be easily bucketed under valuable customer segment.  LegalZoom can choose to ignore the ‘signs’ because from a profitability standpoint my disappointment has no material impact.  But does this mean I’m a customer for life?

Related post: Defection Latency

Customer Segmentation

My dad asked me to buy 3 GPS accessories for him.  He gave me the name of the equipment and the web address of store.  I had never heard of and though it violated quite a few of my ecommerce essentials I knew dad had done his research.  I made the purchase.

Data can be highly subjective.  Looked one way it says something, looked another way something entirely different.  How much money makes off me depends entirely on how accurately they segment me as a customer.  Under one view my transaction was quite standard and I couldn’t fault them for assuming I am a typical site shopper and sending communication messages intended for GPS enthusiasts.  But this view is inaccurate, and a brief review of the transaction proves it:

— This was my first purchase on the site and I bought 3 GPS units.  [Why did I buy 3?]
— I completed my transaction in 32 seconds and spent $150. [This speed of transaction suggests I knew exactly what I wanted]
— I used expedited shipping. [Arguably, I was in a rush]

If I was working with I would have stored these extra bits of data in my customer file.  I would also add a field called “inactivity”.  If Rishi does not open our emails or make another purchase for the whole year it would clearly indicate he is not a GPS (or gadget) enthusiast and made this purchase as a gift for someone else.  Just knowing this could help me make more money through Rishi.  See, they’ve been sending me emails assuming I bought the GPS for myself and this is one reason I don’t read them.  However, if I got an email that said “Hi Rishi, six months ago you bought the GPS 33.  We believe this was a gift purchase and wanted to inform of a new accessory that would be a perfect gift for someone who already has the GPS 232 unit”.  Suddenly I am interested and would love to gift this to my dad.  Context is everything.

Related articles: Capturing Intent and KuKuRuZa Gourmet Popcorn