The Myth of 2-3% ConversionI've noticed a trend in online marketing and usability: whenever people talk about conversion rates, they all seem to quote an average of 2-3%. The problem is that, over time, I've noticed another phenomenon: no one ever seems to back this up with data. Ultimately, 2-3% conversion seems to be a bit like Bigfoot or, at best, a friend-of-a-friend statistic: "my cousin's sister's hairdresser said she read in Business Week that the average website gets 2-3% conversion."
Of course, some sites get 2% or 3% conversion, but is 2-3% the average across all of the commercial sites on the web? I very seriously doubt it, and here's why:
It Depends What You MeasureFrom my own client experiences, I've seen that conversion rates can be very contextual. For instance, if I'm testing two versions of a home-page, a well-designed test will only track the visitors who see the home-page (not the ones who land deeper in the site). For obvious reasons, visitors who never see the home-page aren't relevant to the test. The conversion rates I ultimately see after those tests are often in the neighborhood of 2-3%+, but of course, I'm only looking at a subset of the visitors. I'm also typically tracking visitors with a long-term cookie, which reduces the unique visitor count and may artificially inflate conversion rates.
It Depends Who Measures ItEven if we could look at all of the conversion rates reported and average them, one big problem is that we can only include people who are actually tracking their conversion rates. I suspect that anyone savvy enough to track their conversion is also focused on improving conversion. If we only count the people who know what their number is, we're almost definitely overestimating.
Why Doesn't Google Know?I was surprised at a recent conference to hear a Google engineer quote the 2-3% conversion rate. If anyone could dig into their data (AdWords, in particular) and pull out a reliable average conversion rate, it's Google. Why don't they? Honestly, I suspect the answer isn't as positive as most people want to hear, which leads us to...
High Conversion SellsWould people spend as much as they do on PPC if they thought that average conversion was more like 0.5%? Probably not. In fairness, they probably also wouldn't spend as much on conversion improvement and usability. Having an achievable goal that sounds nice (and 2-3% sounds pretty good to most people) encourages people who are falling short to pursue conversion-dependent products and services. That's not a conspiracy theory, just an economic fact.
How Much is Good Enough?So, what is the average conversion rate? Maybe a better question is: what's a good conversion rate? Honestly, that depends a lot. If you're paying $1/click on PPC, have an impressive 10% conversion rate, but only make $2 in profit per sale, you're still losing your shirt. Meanwhile, if your website drives leads for a $100,000 piece of equipment, even a 0.1% conversion rate could drive impressive revenues.
Consider another question: which of these situations would you rather be in?
- 100 visitors/day converting at 5%
- 500 visitors/day converting at 1%
- 5,000 visitors/day converting at 1%
So, where do we go from here? Well, I think we can start by having more conversations about what helps drive conversion and less about what a "good" or "typical" number is. At the end of the day, your goal is improvement, not chasing some mythical average. If you sell more tomorrow than you did today, you're on the right track.
I think it is healthy to ask these questions with any statistic we see or find.
You are right, I have personally looked into that 2-3% statistic and find EVERYBODY quoting it with no hard data reference.
I enjoyed your last paragraph:
"...start by having more conversations about what helps drive conversion and less about what a "good" or "typical" number is. At the end of the day, your goal is improvement, not chasing some mythical average."
Thanks for the insight!
Dr. Pete· Wednesday, December 19
Thanks, James. I think you're right; we have to challenge any metric we use. It's great that we've moved from being focused on hits to visitors to conversions, but that doesn't mean that the latest-and-greatest can be taken for granted.
I just went and checked my stats for the period Nov 18 - Dec 18. Using Numbers for unique visitors (according to Google Analytics) and actual sales my CR for this period is 2.8% with a cart abandonment rate of 10%. But I agree, it's immaterial, good or bad is determined by relative change over time, not a comparison to another business. I realize BTW that GA has to under report because of several factors.
Dr. Pete· Wednesday, December 19
Stop messing up my point with your inconvenient "facts", Dave! :) All else being equal, I would say that 2.8% is pretty high, so you must be doing something right. I think what concerns me the most is the people who seem to have conversion rate envy and assume they're doing something wrong because they aren't in the 2-3% range that everyone else claims is average.
Sorry about that. Here's a more convenient factoid then - CR and cart abandonment both noticeably improved (I have no numbers handy) after removing a canned "create an account" form from the beginning of the checkout process. A tip I saw in multiple places before implementing.
Another point to take into account is the industry or niche.
A site that specializes in selling roses is obviously going to convert at a much higher rate then any site where price comparison is occurring.
Dr. Pete· Friday, December 21
Mike, you're right: there are a ton of variables that affect CR, and trying to determine some global average across every industry just isn't helpful for most people.
On a side note, I really wonder how some large sites calculate conversion. Look at Coca-Cola, for example; they spend millions on the web, and don't sell their product directly. I'm sure they have a number of secondary goals, like getting people to sign up for promotions, but I'd hate to have to justify the ROI on that website.
Check this post out. I know the director of search at Proflowers -- 30% conversion.
I was shocked when I saw this post.
Then I thought about it and compared to our business. (consumer electronics)
We can set our watch to about 2.5%- 2.75% conversion, but we deal with many visitors that will visit our site, check out the price of a laptop or tv, then run to tigerdirect or newegg and price compare. (same session conversion is very difficult, and rightly so)
With flower shopping you search for a term like "flower delivery" and once on the site, provided the the checkout isn't a nightmare, your done!
Dr. Pete· Friday, December 21
You guys are really trying to prove me wrong, aren't you? :)
I guess it makes sense for certain kinds of retailers, like Proflowers or Office Depot. If I go to one of those sites, I know what they have (they have a strong brand), I have something specific in mind, and the price point is relatively low (i.e. I'm not going to comparison-shop). Of course, they could screw it up with a bad user experience, but the kind of consumers they're getting are the kind that are ripe to buy.
For one of my clients: they're effectively a reseller, many visitors are first-timers, the price-point can be high, and buyers often have to get budget approval (it's a work-related purchase). So, we've felt great about getting their CR above 2%. Sure, I'd love to have a 30% CR, but it's not going to happen no matter how great the usability is.
Agreed, you are 100% correct....
Thanks for joining the discussion over on our blog. Let me address a couple of your points. First the 2-3% average conversion rate comes from a few different sources. The one I typically quote comes from Shop.org that releases their annual benchmark of all their affiliated members. This one is a bit harder to be sure they all are measuring in the exact same way. However, another publicly available source is the FireClick index (here every retailer is being measured with the same software and methodology). Last week, the average global conversion rate was 2.4%. This is also consistent with other non published benchmarks I have seen through out the years such as data from WebSideStory's StatMarket, CoreMetrics Benchmarks and my 10+ years of experience optimizing website conversion rates. I can't agree that this 2-3% is mythical.
I agree with most of what you are saying here in the article. The key is conversion rate is just one benchmark. It should not be the primary metric that is used to judge success. Part of the legend of this metric is its origins in the direct response world where a 2% response was considered successful. At the end of the day, what really matters is revenues. However, a poor conversion rate can be symptomatic of other issues.
Keep up with the good advice.
Dr. Pete· Tuesday, December 25
Thanks for the insights, Bryan. Honestly, my biggest complaint was simply that so many people throw around the 2-3% figure without any data to back it up, and I appreciate that you can bring some data to the table. I suspect that there are still thousands of sites that don't measure conversion and would drag the industry averages down, but it's good to see that more and more major retailers are tracking their performance.
I had this very conversation with a client last week. They wanted to know how they needed to spend on SEO and PPC to get enough traffic to give them 5 conversions a day.
They were somewhat hurt when I came back and recommended that they should look at their site first, fix the problems that stop people converting in the first place, and then worry about getting more traffic.
IMHO it's far more effective if you start off by knowing what your margins are. Then you can work out how many deals you need to cover the business costs. Then you look at your current conversion rates, only then can you start to focus on the real issues.