What is a good landing page experiment’s conversion rate?
Experimenting to figure out what’s effectiveLet’s take the bull by the horns. Running landing pages as experiments requires some thought about conversion rates.
This question comes up in many forms. Here’s a recent post on a forum I participate in. I wanted to address it directly:
Hi!
Is there any benchmark telling that a given conversion rate is good or not good enough to proceed with the idea, getting into next step, interviewing people etc…?
Thanks for your valuable input!
So,
Conversion rates in the context of running an experiment (an MVP) mean something different than on a “normal” landing page. It’s like comparing apples and breadfruit. Both fruit, but quite different.
On a “normal” landing page, conversion rates drive overall profitability. The more people convert (buy), the more revenue you make given the same amount of traffic. The goal here? Be efficient.
Making money is a strong signal you’re onto a good idea. Yet don’t get ahead of yourself, bucko. Massive profits (and a high conversion rate) from cold traffic are usually the result of an a fully optimized sales funnel.
An MVP is a different beast, or bull, to continue the original metaphor.
When running a landing page MVP, in contrast, you’re figuring out what’s effective. What’s worth building.
You use the conversion rate as an indicator of whether your proposed solution will address the problem that you think a particular audience has. Note: lots of guesses in that last sentence.
This is a different approach than “classic” A/B optimizing. You aren’t trying to optimize your idea to convince yourself that you’re so smart (even though you are). You’re trying to figure out whether people actually need what you want to sell. It’s a “go/no go” test. Either you proceed with the idea or you don’t.
The only conversion rate benchmarks which I’ve found in my research are conversion rates in fully optimized funnels. (The first kind).
By current standards, a conversion rate of over 30% to email is fantastic according to kickofflabs.com. Wordstream claims a conversion rate of 10% to a sale is great.
Given that’s the case, you’ll need to decide up front what is acceptable for a a landing page MVP—before you start sending traffic to it. It’s an experiment. You want to learn something whether your experiment crosses the threshold or not, right? That way you get a meaningful result with your hypothesis test.
Stripped down to its core, a landing page MVP is just a sales funnel that hasn’t been optimized yet.
So given the above, here’s how I would approach it:
- a conversion rate of 0% is a clear reject
- a conversion rate of 30% to email and 10% to a low price sale is a strong accept (yes these do exist–I have done a few like that).
The trick is choosing a threshold between 0 and 30% that makes sense for your particular product idea. A fully formed product idea requires an understanding of your target audience, their problem, and a proposed solution in return for something of value. Typically this refers to email or cash. That’s what you need for a landing page MVP.
Given the above, at minimum, you’re best off running a few tests in succession:
- Does the customer have the problem you think?
- Is the customer willing to pay to solve the problem? (you can test this with a low cost offer and charging for a solution)
- What is the optimal price to maximize total profit? (test what the optimal combination of volume, price, and conversion rate are)
This will quantitatively validate whether you’re on the right path, without building a product. It’ll also force you to think about how you’re going to acquire prospects….which is a topic unto itself.
Once you are asking for money on your landing page MVP, it makes sense to simplify the above with the following formula:
gross profit/customer
= revenue/customer – cost/customer acquisition
= price on landing page – cost per ad click / landing page conversion rate
Notice how this relates the price you want to charge for your solution to the target conversion rate you want to achieve. In this context, it’s easier to think about what you need your conversion rate to be for an idea to make sense. As the price you charge goes up, the conversion rate can go down while still making your idea profitable.
If you first offer a high value solution that’s attractively pitched at a low price, you should expect to get a high conversion rate. The higher you make that threshold value, the more ideas you will reject.
So if you apply Wordstream’s 10% conversion rate to a landing page MVP, be prepared to fail a lot of tests in order to find it. It does happen, though.
You’re probably dying to find out what I use. 😀
Currently, I want at least a 5% conversion rate to a sale on cold traffic before I pursue an idea further. If it’s less, the idea is bad, or the audience is bad. In either case, it needs rework. Or a trash bin.
I’ll give you an example of such a failed test tomorrow.
If you are launching a product or evaluating a new business idea, you’ll love my book Launch Tomorrow. It goes into the nuts and bolts of validating and launching new products–using landing page MVPs.
Don’t delude yourself, as I know I have many times in the past.
Get feedback.
From your ideal prospects.
This is a great post, the one thing about conversion rate is that it can sometimes be a vanity metric. For example, if I have a landing page that says, “Pay $1 and get $2 — Sign-Up Now” it would have a very high conversion rate. The irony is that any good product which is priced right essentially gives users $2 for $1, but we have seen this done in a manipulative way, such that the actual offer is impossible to deliver on. ie. “Get a Macbook air for $50” would have a really high conversion rate but wouldn’t be possible in any realistic or scalable way.
Trevor,
Yes, it can be a vanity metric if a conversion isn’t net profitable. The point here is to first figure out what people convert on, before building anything and/or figuring out the economics of product/service delivery. At least that way, you’re only working on products people want.
I’m always a little wary of arbitrary benchmarks but Lean Analytics gives a few for specific industries which I like. The benchmark conversion rate for eCommerce is very different than for a freemium product.
Tristan, yes industry’s also an important variable to keep in mind. Here’s a breakdown by industry of conversion rates I just found:
https://www.marketingsherpa.com/article/chart/average-website-conversion-rates-industry
The key here isn’t the benchmarks themselves.
It’s using the benchmarks to inform your hypothesis going into an experiment. Then, you can make sense of a test outcome (excluding your mom and cat signing up, of course).
Otherwise, you’re stuck with “split testing” vs. a 0% conversion rate, which is hard to interpret.
Very interesting post, thank you very much.
What do you mean by “on cold traffic” in your sentence “at least a 5% conversion rate to a sale on cold traffic” ?
Do you mean without bringing people with services like Adwords ? But then how do you bring enough people in a relatively short time ?
Thank you!!
By cold traffic, I’m referring to any traffic effectively seeing you for the first time. In other words, this will usually be traffic from advertising like adwords for example.
This wouldn’t include visitors who’ve had contact in a non-commercial environment with you before, such as in a guest blog post or content marketing external to your site. That content had already “warmed” them you somewhat, which will increase the conversion rate.
For the purposes of quantitatively testing demand and/or the scale of a problem that’s already pre-existing, it skews your results.