Practical Ecommerce

Quantifying Conversion Funnel Opportunities

In my last article, I discussed the practice of breaking the ecommerce conversion funnel down into four distinct transition points to help establish benchmarks and to locate and isolate the greatest opportunities for profitable improvement. Measuring these transition points is a valuable first step in the process. The next step, however, is where the dollars and cents really come into play.

Here’s a quick recap of the four funnel ratios we’re measuring:

  1. Overall Visits to Decision-Page Visits: The percentage of visits reaching a page with an offer description, pricing, and an add-to-cart option. As noted in my last article, this should be no less than 60 percent.
  2. Decision-Page Visits to Cart Visits: The percentage of decision-page visits taking the next step to the shopping cart. This percentage should be roughly 30 percent.
  3. Cart Visits to Checkout Starts: The percentage of shopping-cart visits taking the next step to start the checkout process. This percentage should be 60 percent, and anything less, typically, warrants a closer look.
  4. Checkout Starts to Completed Orders: The percentage of visits starting the checkout process that actually complete the order. This percentage should be 75 percent, and ratios less than 75 percent at this deepest part of the funnel should prompt a focused diagnostic effort.

Once these four conversion funnel transition points are measured, the gaps between current performance and expected benchmarks at each transition point can be quantified — effectively turning the percentage differences into revenue- and profit-value estimates.

By quantifying the performance gaps at each stage, the value of making improvements becomes clearer and much more compelling. Let’s face it, percentages aren’t very motivating, but hard dollars can inspire massive action! Quantifying performance gaps also helps with prioritization and budgeting — ensuring a focus on the most significant opportunities and maintaining appropriate investment levels for improvement efforts.

The Quantification Process

Quantifying performance gaps is a relatively simple process, but it does require some math. And it also requires some additional information about your typical order sizes and margins. To illustrate the process, I’ll use an example case from my client files.

First, we nail down baseline performance and calculate the magnitude of the performance gaps relative to expected benchmarks.

From here, it’s much clearer where this company’s greatest opportunities for profitable improvement are located. In two areas—decision-page visits and checkout-start visits — the current ratios exceed the benchmarks. At the same time, shopping-cart visits and completed orders are currently performing well-under the benchmarks. In this case, it would be a better use of resources to shore up the poor-performing areas, rather than attempt to squeeze even more out of the areas that are already performing pretty well.

But just looking at the percentages can be misleading because the largest numbers don’t always represent biggest opportunities in terms of revenues and profit.

To estimate the revenue and profit impacts of improving performance to benchmark levels, we use the firm’s typical order sizes and margins to quantify the negative gaps while holding the other funnel ratios constant. In the shopping-cart visits area, for example, we estimate the incremental value of the performance gap by recalculating the overall funnel, substituting the 13 percent gap for the 17 percent current shopping-cart ratio.

To estimate incremental revenue in the shopping-cart area, the formula would look like this:

32,500 x 68% x 13% x 64% x 48% x $114 = $100,615

Similarly, to estimate incremental revenue in the completed-order area, the formula would look like this:

32,500 x 68% x 17% x 64% x 27% x $114 = $74,010

We now have a better estimate of the value these performance gaps represent in terms of dollars. Isn’t $100,615 a much more motivating figure than 13 percent? You bet it is. And if we really want to get excited, just remember that these numbers are weekly—we can annualize our estimates by multiplying by 52 for some eye-popping impacts.

Understanding the Quantification Results

For Company “X”, focusing their limited resources on improving their decision-pages and checkout process are clearly highly-profitable activities. Of course, the task of diagnosing the specific problems and identifying root-causes in each area still remains. But this further diagnostic process is a whole lot easier when focused on specific areas.

Should the firm tackle both areas at once or individually? It depends on the firm’s available resources and ambitions. But I usually recommend making improvements in more than one area at a time. Why? A conversion funnel is an inter-connected system where improvement impacts combine and compound as they move through the rest of the funnel. In this way, improvements in multiple areas can produce compound impacts far greater than the sum of their parts.

By focusing on specific areas, does that imply that there’s no room for improvement in the other areas? Certainly not — even where benchmarks are being surpassed, there is always room for improvement. It’s just that in my experience, it’s much more difficult to improve something that’s already performing very well. I’ve found that it’s usually far easier and more efficient to identify and implement big-bang optimizations in areas with significant performance gaps.

It’s important to remember that this quantification process produces estimates. There are a number of variables that aren’t accounted for in this simplified model — differences in order sizes and margins, changes in other funnel ratios, and variable transaction costs, to name a few. Of course, you can choose to account for some of these elements when creating a quantification model specific to your situation.

Practical Ecommerce

Practical Ecommerce

Bio   •   RSS Feed


Sign up for our email newsletter

  1. Legacy User September 13, 2007 Reply

    Thank you, the concrete application examples are extremely helpful.

    — **

  2. Legacy User September 14, 2007 Reply

    Wouldn't the number of products that a retailer (for example) carries be a determining factor of conversion rate if you were using product categories as keywords? For instance, one of my clients is considering hiring an SEO company to help rank 15 keywords/phrases.

    They'll be using product categories as the keywords/phrases. The SEO company charges a monthly rate and guarantees that each keyword/phrase will be in the top 3 pages of one of the 3 search engines (Yahoo, MSN, and Google).

    Is there a way to guesstimate how many products per category they would need in order to (at first) just cover the monthly rate?

    — *Wes*

  3. Legacy User September 16, 2007 Reply

    Hi Rafé,

    Great posts!
    Just a short question on how to compare the 8% order to conversion rate (100*0,6*0,3*0,6*0,75) with the 2%-4% conversion rate that is generally taken as average?

    — *Béate Vervaecke*

  4. Legacy User September 15, 2007 Reply

    Hi Wes,

    While not directly related to conversion funnel analysis, I think I understand where you're going with your question…

    Without knowing how demand breaks out between specific products within a product category, it's extremely difficult to know how the company's product spread is going to affect the performance of SEO efforts.

    But it's certainly possible to understand the results that are required just to cover the expense — i.e. calculating the amount of business, orders, or customers that will need to be generated to break-even on the incremental expense, given the company's margins and transaction costs.

    From here, qualitative judgements can help gauge the likelihood of achieving those results. For example, if the product category keywords are very general and the company is really only competitive on a handful of products within that category, it's likely that there'd be a tremendous amount of waste in the campaign. If, however, the company's offerings align well with the breadth of the product category keywords, waste will be lower.

    Obviously, the greater the alignment between the keywords being ranked and the actual offerings of the company, the greater the likelihood of achieving an acceptable ROI.

    Hope it helps.

    Rafe VanDenBerg

    — *Rafe VanDenBerg*

  5. Legacy User September 18, 2007 Reply

    Hello Beate,

    To answer your question, the reported industry averages are, well, averages. And the lower-end of the distribution really pulls those numbers down in a big way. The benchmarks we use for diagnostic purposes are based on our optimization work and they strive to balance conversion efficiency across the funnel with profitability considerations.

    Granted, our clients are likely very motivated and probably not reflective of the average in this respect. But for those interested in above-average performance, these benchmarks are achievable aims.

    Is it possible to go much higher? Certainly. There are more and more e-commerce companies focusing hard on their conversion funnels and achieving very high ratios as a result. But in our experience, higher is not always better.

    When we see overall conversion going much north of 10-12% for a company with relatively "normal" margins, that's a big red-flag for us — it's usually a sign of sub-optimal pricing or targeting criteria that's far too tight — either way, we usually find that profit-dollar gains are being sacrificed for those really high ratios.

    We strive for the most-profitable conversion ratios — and not necessarily the highest ratios — and that's reflected in our diagnostic benchmarks.

    Hope it helps.

    Rafe VanDenBerg

    — *Rafe VanDenBerg*

  6. Legacy User October 22, 2007 Reply

    I don't agree with having hard and fast benchmarks as a catch-all for everyone's cart health. It is a very slippery slope. The author states that checkout "Starts-to-Completed Orders" ratios under 75% are a big problem. That may be true if you are selling a $2.00 item, but how about more considered purchases like a TV or computer? Your cart completion ratio will probably be much smaller because people need to shop, do their research, and then buy. Also, while the funnel does provide an easily understandable view of your cart, it doesn't take into acount anyone who isn't shopping in a linear fashion (most of us). Don't get me wrong, I appreciate commentary on how cart funnels can be a part of the analytic picture, but I worry when it is used as the end all source for how your cart is performing.

    — *Justin*

  7. Legacy User October 29, 2007 Reply

    Hi Justin,

    The diagnostic process we're talking about here shouldn't be taken to imply a linear shopping process. The funnel points we measure are simply milestones that must be passed through in order to complete a purchase. Regardless of all the bouncing around a shopper will most certainly do, they have to pass through certain milestones.

    Whether or not prospects are reaching these milestones in adequate ratios is what we're looking for — excessive fallout in different areas of the funnel points to different underlying strategic and tactical root-causes.

    Shoppers who reach the lowest portion of the funnel, regardless of purchase price, have self-qualified to much greater degree than general site visitors and have demonstrated a high-degree of purchase intent — more than 25% fallout at this lowest level should indeed raise eyebrows.

    Rafe VanDenBerg

    — *Rafe VanDenBerg*

  8. Legacy User October 31, 2007 Reply

    Rafe, Thank you for replying to my message. I very much agree with you regarding the point that milestones in the cart and elsewhere on the site need to be monitored to determine where to "raise eyebrows". My trouble lies in the term "excessive fallout".

    I am still questioning the 75% cart-starts-to-completed-orders info. According to Marketing Sherpa, the industry average for cart abandonment is 59.8%. How then is >25% fallout a "raised eyebrow moment"?

    Thanks again.

    — *Justin*

  9. Legacy User November 9, 2007 Reply

    Hi Justin — Good questions, but I think we're talking about different things…

    Checkout-starts-to-completed-orders is a different ratio than the cart abandonment ratios cited in industry reports. Whereas cart abandonment is just the ratio of carts that don't complete an order, the "checkout start" is further down the funnel — measuring not just carts, but the carts that actually begin the checkout process.

    In other words, we break the cart abandonment ratio into two ratios — cart-to-checkout-start (60%) and checkout-start-to-completed-order (75%). This allows us to see where the real issue is — a problem in the cart-to-checkout-start ratio points to very different root-causes than a problem in the checkout-start-to-completed order ratio. When taken together, these two ratios are somewhat similar to reported industry averages for cart abandonment.

    But keep in mind that industry averages include both the good and the bad. And as an e-commerce optimization firm, we help our clients do better than average — so it shouldn't be surprising that we're able to set the bar a bit higher.

    To sum up, a >25% fallout for checkout-start-to-completed order should raise eyebrows because this is the deepest portion of the funnel, representing the greatest level of qualification and purchase intent. When we see sub-par ratios at this point in the funnel, we begin looking for problems with required registration, slow SSL response, poor form design, payment and shipping options, etc.

    Hope it helps clear things up.

    Rafe VanDenBerg

    — *Rafe VanDenBerg*

  10. Krystofar November 26, 2012 Reply

    Where can I find these four primary measurements in Google Analytics?