ecommerce site abandonment Options

[cre8pc]

Am typing with an injured hand... High winds took the kids trampoline - one of those big ones with the net - up over our fenced in backyard and into the neighbors yard. It's sitting in tangled heap over part of the now broken split rail fence and tall evergreens...deep cut on my hand while a bunch of us were trying to move it (it's very big) in a strong wind and driving rain.

Anyway, thanks to Alan Webb for this find.

Closing the Deal Online

It is important to realize that while customers often don’t know what they want, they do know when they get frustrated or confused.

[James]

Hi Kim,

Nice find. I found the analogy with the local grocery store interesting.

Of course, the difference is that with the local grocery store:

(a) there is probably a small range of competition in the local area
( trust is very quickly built up as you can see the range of products, quality as well as the professionalism of the organisation, staff, etc
© the relationship between store and customer has probably been built up for many months or years
(d) local stores tend to be a well-known name, you drive past them everyday, see their advertising on the television, in the papers, etc, etc
(e) the simple fact that there are hundreds of cars in the car park leads to a credibility advantage

With an internet store:

(a) the competition is far greater, particularly with geographical boundaries fast eroding
( we are left guessing at the quality and credibility based on what we read and look at on the site (as well as 3rd party opinions)
© customer loyalty on ecommerce sites is soon swayed by cheaper prices, experience of poor customer service, etc, etc
(d) a domain name can soon be forgotten given the number of competitors
(e) most ecommerce stores don't advertise how full their car-parks are let along how many are heading through the checkout

An interesting conversion rate to measure would be:

how many web users who start out with the good intention of purchasing their product online end up purchasing online?

[TheManBehindTheCurtain]

The author touches on the real question about conversion (to me, at least) at the end of the article when he begins to acknowledge that we must factor the customer's intention into conversion metrics:

Without understanding the intent of the customers, the retailer might assume the merchandising qualities of the page are insufficient. This may be true, but it is also possible that there is a large customer base that has no intent in buying online, but rather uses the Web channel as an effective research tool to make store visits more efficient and productive.

Most people who pull into the parking lot of the local market aren't there to check the nutrition information on the Jello package. They're there to buy Jello. My guess is a very small proportion of people walk into a grocer to browse. On the other hand, how many mall-goers wander into Macy's just to window shop, with no intention of buying. The question of intention differs by location and by store type.

I visit Amazon several times a week to read the customer reviews of books and other products. But I buy there only maybe once a month. My visit-to-purchase ratio is relatively low. But here's the important bit of information: I don't buy books online from any other source. Ever. In fact, I only buy books from local stores when I need them faster than Amazon can get them to me with their free shipping and no sales tax. Or I'm picking up something from the sales rack. On other sites -- say, ordering flowers for delivery -- my visit-to-purchase ratio is very high because I'd never think to go there under any other circumstance.

Until we're able to deeply examine the top end of the funnel to sort out the actual shoppers from the browsers, discussion of conversion metrics may be based on skewed assumptions. For example, a high abandonment rate may not be all that bad if you can prove that the high traffic shows your site also has credibility as an information portal. While your per-visit conversion rate may look low, the effect of larger numbers of people visiting, even while buying at a "lower than expected" rate, may still amount to a valid business model. Trying to drive browsers to conversion may be the wrong strategy if it succeeds in diluting the trust that visitors have in you as an information portal. Maybe you get higher conversion, but far less visitorship overall, and business actually suffers.

Well, it all goes to show that this is a very complicated business, and that we're in the infancy of trying to understand how to build business models around it.

[DianeV]

I am always curious about this method of looking at stats. In short, stats-in-a-vaccuum are often not a true measure of what's happening. Frankly, I find bean-counting often dreadfully naive and devoid of ... well, life.

Recent examples:

-- I visited the Micron website (buympc.com) several times over the last few months; played with their computer configuration tool a number of times; emailed the configured specs to myself; added it to the shopping cart a time or two; dumped it all and left. More than once. Finally, I called Micron directly to get answers to a couple of questions, got a salesperson who I'd say was really a tech, and configured a somewhat different (and more expensive) machine that arrived precisely seven days after order (earlier than they promised), prior to Christmas, at the busiest time of year. Is this an example of the "shopping cart abandonment rate"?

-- Decided I wanted a computer Mahjong game (bad idea; too engrossing); found several in Google; downloaded three trial versions; liked one. It timed out after 20 games. I went to the website, added it to the shopping cart; realized it was some foreign company I'd never heard of, got nervous and dumped it all and left. Shortly thereafter, decided I couldn't live without "my" Mahjong game; visited the site again, realized the cart was via Regnow.com, called Regnow.com to find out whether they provide the credit card numbers to their software makers; answer was no; I purchased it online.

-- I occasionally make small purchases of nutrients from a company online even though their shopping cart is ridiculously difficult to use (a "please enable Javascript" notation would solve all). I've sometimes priced a selection of items by adding them to the shopping cart, and abandoned it only to purchase later. This year, we've purchased 10 times what we've purchased in any previous year.

Sum total: many thousands of dollars in sales, every one of which involved at least one abandoned shopping cart. According to the "abandoned shopping cart" articles, the above companies are missing out big time because their shopping carts were all abandoned by me. I submit, instead, that this theory falls prey to one-channel marketing vision which leads to false statistics analysis and poor conclusions. (Love those sound bites, though.)

In particular, I find these two ideas to be false data:

(1) The idea that everyone who puts anything in a shopping cart is really ready to buy.

(2) The idea that any abandoned shopping cart means you won't make the sale or, even, that there's necessarily anything wrong at all. People don't always part with their money so easily or right now or at all just because you want them to.

There's a little more to sales than that, which includes dealing with real, live humans rather than a bunch of units moving through a website. Bottom line: if you're laboring under the idea that people are going to swoop into your website, get dazzled by it all and make a purchase first time and every time without ever questioning anything or thinking it over or postponing 'til later or until they get their tax refunds, then I guess you can live with "shopping cart abandonment rate" echoing in your ears. <wink>

The local grocery store is a bad example; people go to grocery stores more often to buy than to window shop or to price items; they do, however, window shop at clothing stores on a regular basis. The fact that they don't buy immediately in one smooth pass from the door through the store to the checkout counter is not indicative of anything. And yet, that seems to be what is expected of websites, with the conclusion that, if it doesn't happen just that way, you're failing. Nope, that isn't real life, and people don't have to be "confused" or "frustrated" by anything at all in order not to buy right now.

This is why it's important not to look at stats in a vaccuum. What's important is to get overall sales figures up.

<Edited to clarify. /Diane>

[DianeV]

Here's an Internet Retailer article, Holiday web2store shopping rises 43%, CrossMedia reports, that notes:

The number of holiday season shoppers who browsed its client retailers’ web sites before purchasing in stores rose 43% this year over the 2003 holiday season, to 14.7 million from 10.3 million, CrossMedia Services Inc. reports.

And why not, if customers want to ensure that they receive their purchases by a certain date?

That makes sense. If one is looking for a specific item and lives, as we do, in an area where many stores *might* offer that item, I'd far rather determine availability and pricing online before driving all over town ... which I might do if the item is available in a local store(s), meaning that I can get it today rather than waiting for shipping.

Unfortunately, this can easily mean that the effectiveness of a website is difficult to gauge, as it is not credited with sales it helped to make. It may mean that the website actually convinced me to buy, but that I used the bricks and mortar store as my alternate delivery method.

This is why I like to look at the overall sales and marketing figures, rather than just Web stats. It may just be that the work you are doing is more effective than the stats report, and than you think. I've redesigned sites for which the client(s) far preferred to have people call instead of filling out an online email form; when the quantity and quality of calls rose, we knew we'd done our job. A simple "how did you hear of us?" question confirms that.

<More editing to clarify. /DianeV>

[bragadocchio]

Nice posts, Diane.

I like the description of four main types of visitors that Andrew Chak describes in his book Submit now (Review here - which roughly describes them), but I think that we do need to add at least a fifth type of visitor to his descriptions - people who use a combination of web research, and actual offline contact to make a purchase.

Diane's examples of situations where that can happen are excellent.

They are also a reason why I like to include mailing address, email address, and phone numbers on every page, if possible. It really doesn't matter if a person orders online, or if they call or email and ask questions and then decide to make a purchase.

I've redesigned sites for which the client(s) far preferred to have people call instead of filling out an online email form; when the quantity and quality of calls rose, we knew we'd done our job. A simple \"how did you hear of us?\" question confirms that.

It also can help to pay attention to the types of questions people have when they call, and to figure out if you can address some of those on the web site as you go. It really is wonderful when the questions are informed ones because your site helped educate your potential clients.

[DianeV]

> It also can help to pay attention to the types of questions people have when they call, and to figure out if you can address some of those on the web site as you go.

True. It's even more vital to determine which questions you should answer on the website. <grin>

[cre8pc]

There's no doubt that abandonment issues are no longer obvious or reasons easily detected or known, which is why user testing is becoming so important. And not just any "users" either.

How I proceed to navigate a motorcycle site is going to be way different than how someone who lives, breathes and eats "hogs" is going to manuver around it. For starters, they'll understand the terms, they'll know what they're looking for, they'll bring their experience with them...vs me, who knows nothing and will likely simply stare stupidly at the pretty pictures of shiney new bikes and not understand anything written in the description. My leaving the site would be because I'm not the target buyer.

Soooo, stats are stats and will overlook details. Ecom sites will have to find more and more ways to get better feedback while visitors are on the site, or find ways to follow up. And, expectations for shopping carts have to change. We believe people put things into them because they intend to purchase them. But the Internet is not Wal-Mart. The web is a huge store containing the single store which you found that has a thousand other competing stores bumped up right next to it, not down the road or in the next town, all presenting the very same item.

Of the five or so stores I purchased from for Christmas shopping, two of them are following up with surveys after the shipment arrives, one store sent me a coupon for my next purchase, all of them stayed in constant contact with me until they shipped their merchandise and one just emailed wanting some feedback on one of their products specifically.

Of the five, I remember that for three of them I put things in and out of the shopping cart. Sometimes I'd find a better deal on their own site, or would realize I ordered the wrong size and had to start over. There's no way for me to communicate to their software my thinking as I go along, and they're left with just stats.

[DianeV]

Exactly, and that was my point, Kim.

The stats with regard to ecom sites, and particularly shopping carts, don't take into account how people really shop; instead, they assume problems that may or may not exist and use scare tactics to imply that site owners don't know what's happening and are losing out (but of course, the article writers know, and should be hired).

I'm not suggesting that there is nothing wrong with any ecom site or shopping cart. I am suggesting that more has to be known about any set of stats, and the real usage occurring at any site and reasons behind that usage, before anything is assumed. Particularly with articles and analyses that apply blanket assumptions across the entire Web.

[kensplace]

Have you had your hand checked out medically? Dont want to risk any long term damage - typing is good Damage is not!

[cre8pc]

Actually no, but it was determined by two who saw it (husband and neighbor who is a fireman) that it had needed stitches. We were in the midst of a wind storm and objects were flying around the property, including the large trampoline that jumped our backyard fence and broke part of the fence. I was focused on keeping my kids and my neighbors kids calm. My daughter helped me tend to the hand. She was quite brave about that (she's 14.)

The hand is healing but I'll have a scar for sure Thanks for your concern!

Diane, yes I knew that was your point. And a darned good one too! Frank's post is great too...good examples there.

[kensplace]

Ouch!, ah well, scars can be sexy

Sounds very windy over there, we had high winds recently, but nothing like what you described.

[cre8pc]

It was surprise really. I live on a ridge, with a valley below where most of the little town sits. The ridge is actually the very start of the Pocono Mountains. From here on north, it gets hilly and then you hit mountains (though small ones at first.) Very pretty, but it gets a little windier here than one would expect.

Seeing a flying trampoline is very strange indeed.