Data mining from customers who did not buy.

Using a database of abandon carts to survey customers

[lgn1]

A few months ago, we started logging customers who started the checkout process and have entered address, email and phone number information, but who have latter abandoned the cart before completion of the sale.

Our plan is to conduct a survey, to find out any weak links (ie prices to high, shipping to high, to small of a selection, unconfortable buying on web etc.)

I can either conduct the survey by phone or by email.

I feel that an email survey will be treated like spam. Like how many people actually respond to those Dell customer service surveys by email, even when you buy from then.

However, people are sick of telemarketers, pollsters and the rest of the unsolicted calls.

So im undecided, which way to proceed.

Any suggestions or experience with this?

[shri]

I was in a discussion with some folks who specialize in this sort of stuff.

They use a call center in Asia to call your abandoned leads / sales and try to close them for you, at a cost.

Seemed to work for the people who were involved in this deal. It was not ecommerce related.. was related to lead generation, but I figure this approach would work in both cases.

Based on what I heard, I'd recommend a personal call saying "Hi, I'm calling from site.com and we noticed you had placed an order for widget1 and 2, but did not complete the order. Is there something we can help you with? Yes / No -- Then help them or ask them why they did not complete the order".

[PhilipM]

"However, people are sick of telemarketers, pollsters and the rest of the unsolicted calls."

You have answered your own question.

[lorax]

Welcome to WebmasterWorld PhilipM!

I often wonder why folks sweat the details of cart abandonment. Don't get me wrong. I think it's important to make sure the cart and checkout are built as well as they can be but once that's been done, I think it's better to work on tweaking the conversion from the other end - bringing in the right type of customer.

I simply don't think it's worth the all of the headaches and worrying over whether or not you've got the right color in the right place at the right time. I may be too old fashioned but I prefer my ecomm clients not worry so much over these details and would rather they focus on developing long term customer relationships and reputation. To me, these will provide far more sales over the long term with less aggrivation.

[iamlost]

Several years ago a client had lgn1's concerns so we ran a multiple approach:
* we logged site-path histories and points of disconnect plus data uploaded.
* we added a pop-up (this was back when they were merely irritating not actually detested) querying (checkboxes) why they were leaving.
* we phoned a random selection of those who had provided a number offering a free gift in return for answering our questions. Note: we were able to 'check' some answers against those provided via the pop-up.

The results were interesting.
* Most that replied via the pop-up likely lied.
* Almost all that were phoned were surprised that we had their number - they had thought that abandoning the checkout wiped previously submitted info.
* Over 60% that were phoned were mad at us for 'using' their information.
* Barely 50% that were phoned agreed to answer our questions and the results were generally uselessly vague. Over half apparently didn't remember visiting the site at all.
* graphing the site-path to disconnect data pointed out several 'points of concern' which led to increasing simplicity and availability of information (not an easy task :-)).

End result:
The completion percentages did go up noticably if not dramatically. We did receive a few nice comments on 'how easy' buying from the site was. They still log and graph looking for things to improve.

Note: often the easiest way to decrease cart dropouts is to either include 'free' shipping in the price or to show shipping cost alongside item cost. This way many can leave before queuing and just be browsers rather than 'lost' sales.

Finally: I agree with lorax.

[lgn1]

Thanks for the response and the experience.

I am now more inclined to go the Email route.

Although, my response rate will be much lower, it will be
much easier to manage, once setup, and I can run the campaign over a longer period of time, with little additional cost.

Some people may consider a site survey as spam, but it is alot less likely to raise blood presure, by hitting the delete message, rather than answering, explaining or swearing and then hanging up a phone.

Any who knows, some of those customers may come back and actually buy, if we offer a carrot.

[john_k]

Note: often the easiest way to decrease cart dropouts is to either include 'free' shipping in the price or to show shipping cost alongside item cost. This way many can leave before queuing and just be browsers rather than 'lost' sales.

You might want to consider this a little more strongly. Is there some piece of information about the final/bottom-line price, shipping, etc. that the potential customer cannot get to without providing their email address, name, and/or phone number? Any at all? If the answer is "yes", then you will likely find that a great number of people proceeded as far as they did just to get to that information. Maybe they never intended to buy at all. But if your site is designed to require account creation before a customer can see all of the costs and details, then you are inviting people to create abandoned carts. In this case, I would guess that a high percentage of people receiving your survey email will see it as a negative.

You cannot discount the fact that many people in the "fact-gathering" or "pricing" stage of the sales cycle will load up carts with no intent of making a purchase that day. Some of these people will normally return when they are ready to buy. If you give these people a negative reminder about your company, they may return to someone else's site to make their purchase.

I estimate that about 15% of my abandoned carts are from people that return 1 or more days later to make a purchase. (this from reviewing IP addresses and items of abandoned carts and finished orders)

[BananaFish]

It would probably better to improve your conversions by improving usability and optimizing other on-site elements. A lot of customers don't buy at first but come back later to purchase, popups, emails, etc may "turn the worm".