11 - Calculating Conversion
Ultimately, the goal of marketing is conversion: getting the attention of as many potential customers as possible is good, but unless you can convert some of them from potential customers to paying ones, it's wasted effort.
Also, there are various steps along the way, and you can measure success as the achievement of any goal: if you can get a person to show positive interest that might lead to a sale (sign up fro a seminar, download a brochure, etc.), that may be as close to conversion as a Web site can get.
Qualification Process
A milestone on the way to conversion is qualification: a person who has shown interest in your product is a prospect. Only those who have the desire and ability to actually make the buy is a qualified prospect. Differentiating one from the other is key to spending your resources wisely.
Generally, there are steps on the way to a purchase:
- Acquisition: The company seeks out prospects, advertising to gather their attention
- Information-Seeking: A prospect is proactive in seeking information about the product
- Negotiation: the prospect engages the seller in a negotiation t get commitments on product specifications, payment, delivery, etc.
- Agreement: the prospect agrees to the purchase
- Settlement: the prospect pays for the purchase
Generally, this is a "funnel" with a large number at the start, and a small number at the end, with drop-outs along the way. The shape of the funnel tells a company where its marketing is effective:
- If the funnel looks like a trumped (flared top), the company is getting the interest of a lot of people, but few of them are actually interested in buying
- If the funnel looks more like a champagne glass, the company is getting a lot of attention from those who are interested in buying, but cannot commit (common for luxury goods: everybody wants it, but cannot afford it)
- If the funnel looks like a shot glass, with smooth sides and the bottom only slightly smaller than the top, this means your marketing is right on target: you are not wasting dollars on unqualified customers, and are closing the deal with those you reach.
- If the funnel looks like a zombie glass (straight sides), then your marketing is perfectly precise (everyone you advertise to becomes a buyer), but chances are you are missing potential sales.
Navigation and Search Impact Conversion
An early step in success is assisting the customer in finding product information.
This is done externally, by making sure that customers who are using search engines to find information about a product are being directed to your site. It is managed internally by making sure that the information is on your site to be found, and that the user can navigate to it from the page on which they landed.
A direct path is not always the best option: some shoppers are browsers rather than hunters, and they will look at a product, then browse about the site rather than charging directly from the shelf to the cash register.
An interesting study: when looking at total sales dollars on a general merchandise site, it was found that only 13% of products purchased were found by using the search engine. The vast majority (87%) were found by people browsing through the categories. It is also suggested that people who browse, rather than search, make larger purchases in the end (more items, more dollars), though statistical proof is not presented.
Another study at a music retailer: when the company loosened search logic, enabling customers to see near matches to what they were looking for rather than the precise thing, they bought more often. Precise figures were not available, but a company that develops search for product catalogs indicated that the increase was between 50% and 200% conversion rate.
Duration
The jury is out on whether the duration of a visit (time spent on the site) is a good thing or a bad thing. There does not seem to be a statistical correlation between duration of visit (time spent) and the purchase volume.
In some businesses (subscription sites), duration is actually a bad thing: if customers pay a fixed priced for all-you-can-eat access, longer durations mean more expense without greater revenue. However, users whose duration on the site decreases are likely to drop their subscription, so it's a careful balance.
It is suggested that more research is necessary. Rather than looking at the duration of the visit, consider which pages the user spent time viewing. A user who spends ten minutes on product pages may buy more than a person who spends ten minutes looking at the shipment alternatives. Breaking down an amalgam into components can be more telling.
Depth
The "depth" is differentiated from duration in that it measures the number of pages viewed rather than clock time.
No statistics are presented to indicate the importance of depth, and it's implied that it's important to consider categories of content rather than content as an aggregate to determine the value of depth.
Recency and Frequency
A common misconception is that a prospect's decision to purchase and the actual act of making the purchase happen in a single Web site visit. In truth, a customer may visit a site several times to research product information before coming back to buy - and when they return to buy, they may not need to peruse the product information, having seen it in previous visits.
Frequency can be telling: how many times did a person visit the site in the week/month/year before making a purchase?
Recency is also mentioned: how much time elapsed between the last non-purchase visit and the visit at which the purchase was made? My sense is that this is more important to marketing efforts (sending a reminder promotion to someone who is on the cusp) rather than performance metrics.
A conversion model (Moe/Fader) has been set up based on four components. A lot of effort is put into describing the components, but the actual algorithm isn't disclosed.
Abandonment
A phenomenon that befuddled and panicked online retailers for years is shopping card abandonment. Studies show that a large percentage of users who visit a site and begin assembling a list of items to purchase later decide to bugger off and not buy anything at all.
It was assumed that these were instances where sales were lost at the last moment, and that the company is doing something dreadfully wrong that is causing customers to ditch. Studies later showed that this is not the case. People who abandon carts often had no intention of making a purchase: they were merely browsing, or putting items in the cart as a method of comparing alternatives.
A better measure of abandonment is the number of individuals who initiate the purchase process (they click "check out") and then abandon - this is clearly a sign of problems.
One independent study of customers who abandoned carts indicated the reasons they did so (check all that apply, so the percentages don't' add up):
- Add-on costs (tax, shipping, etc.) too high (72%)
- Just browsing (61%)
- Changed mind (56%)
- Total cost of items too high (43%)
- Checkout requires too much personal information (35%)
- Site requires registration to purchase (34%)
- Concerns about the security of checkout process (31%)
- Checkout process too confusing (27%)
Also worth noting: a lot of work was put into persisting abandoned carts, so that a user who returned later could pick up where they left off. In practice, this didn't work out so well: an infinitesimal number of users actually bought items that were previously added to a cart; most simply abandoned the carts; and a few emptied the carts and started over.
After the Sale
There is a common oversight among many e-commerce managers that the sale is the end of the road, and little consideration is given to those who have bought. Sites that take the attitude of "once we have your money, we're done with you" often find that they have a lot of shoppers who buy once, and never again.
Successful sites (amazon, monster, etc.) take into consideration that a sale is the first in a chain, and that there is much to be done to retain the loyalty of a customer so that they come back to the same source, the next time they want to buy.
A few case-studies are provided that show how retailers analyze sales data with an eye toward winning repeat business. It seems to vary greatly from shop to shop, and I don't see anything universal.
Channel Conflict
Companies that sell through sales reps have long seen the Web as a challenge, and have often faced great internal resistance from sales reps who felt that the Web was stealing their commissions. The same is said, to some degree, of franchise operations and manufacturers, whose online sales pout them in direct competition with the retailers who stock their goods.
The author provides a few case studies that show how the Web was used as a method of generating leads for salesmen, or referring customers to a retailer, rather than doing direct sales.
The Web can also provide a value-add by providing follow-on support to customers, which decreases the number of product returns and support calls to sales reps who would rather be spending their time getting additional orders.
There is no reliable method for following a lead through the channel, to ascertain that the leads generated via the Web site led to actual sales, or that the support information actually led to decreased returns and support calls. This information can be estimated by various means, but not directly measured.