How often does this happen:
A company budgets for a website, it’s built according to the budget and then it’s done. Hooray! … and everybody moves on to other important things.
Unfortunately too often.
Table of contents
Declaring it done is a waste of money
Your website is not done. Declaring it done is arrogant, stupid and will cost you a lot of money.
A newly built website is a hypothesis at best (hopefully it was designed with the user in mind). Now the real world test starts – and with it the process of continuous optimization.
No Plan Survives First Contact With Customers
If you just allocated a budget for launching a new site, you basically paid for a rough diamond. It works, but it doesn’t sparkle. A cut and polished diamond is worth so much more.
It’s impossible to know in advance what will work the best for your customers. Yes there are heuristics and best practices, but they are a mere starting point for optimization. Once your new site is launched, you need to start optimizing it to figure out what works the best. Optimizing is the best thing you can do for growing your revenue.
A mindset change is needed
The era of just building websites, and keeping them the way they are until the next “new website” project, is gone. If you’re stuck in that mindset, your business is going to shrink and your smart competitors are going to eat your lunch (and dinner).
Your website budget has to include continuous optimization – it’s a non-removable part of your website. Having a website is not a goal – it’s a means to an end – revenue. Optimization is about getting MORE revenue out of your website (improved marketing efficiency).
It takes (depending on your traffic volume) at least 6 to 12 months to pick the low hanging fruits and get the initial gains (often very significant amounts of extra absolute dollars earned). But it doesn’t end there – it’s a non-stop process. When the conversion industry creators, Eisenberg brothers, work with a client, they have a 3-year perspective in mind.
My friend Chris Goward advocates what he calls evolutionary web design – ditching radical makeovers at all and just replacing it with continuous improvement. I support that idea wholeheartedly (with some exceptions – there ARE quite a lot of awful websites out there that have nothing going for them, those need a fresh start to get moving + continuous optimization).
It doesn’t matter if you’ve been doing it for a long time. You may *think* you know what people want and how they want it. However, you don’t – you must talk to people, understand them and run experiments.
For you to achieve your goals, visitors must first achieve theirs.
– Bryan Eisenberg
Data, Not Intuition and HiPPOs
Most sites are still built the stupid way – based on a committee meeting (where none are conversion analysts) or the HiPPO (highest paid person’s opinion).
Another prominent approach is copying what your competitors are doing. Trust me, most of them don’t know what they’re doing, and their website looks the way is does due to a combination of random factors. Plus, they probably copied somebody else.
That’s a surefire way to crappy results, but even that could be salvaged and turned into something that works through experimentation and continuous improvement.
Continuous optimization is based on actual user behavior data. “But we used data when designing the site!” Great! We should use behavior data every time we design a site! The problem is that even when your new site is designed by solely relying on user behavior data, it is a design based on data that happened in the past, on the old site. Your new website is a great hypothesis at best.
Better than the average person, but that’s not enough
Every conversion consultant has heard this a million times: “So what will our conversion rate be after you work on our site?” If someone indicates they can answer that, they’re lying. Nobody can.
My friend Craig Sullivan says that even though he’s run thousands of tests over the years, and he’s better at guessing the outcome of the test than an average person – he’s still not that much better than flipping a coin. And this guess is made when running experiments based on solid data! Experienced optimizers know that you can’t be sure of anything – especially when you’re designing a new site.
Your success depends on the number of experiments you run and the speed of implementation. Knowing what to look for, how to operate with the data and using a structured approach will ensure your experiments are done right and they actually have a chance of achieving significant lifts. Traffic and time are precious resources that shouldn’t be wasted on random tests.
Marketers still rely on intuition
A study of nearly 800 marketers at Fortune 1000 companies found the vast majority of marketers still rely too much on intuition – marketers depend on data for just 11 percent of all decisions.
Funny that almost 100 years after the book Scientific Advertising came out we’re still relying on intuition more than anything. Dan Ariely notes in this excellent article:
Companies pay amazing amounts of money to get answers from consultants with overdeveloped confidence in their own intuition. Managers rely on focus groups—a dozen people riffing on something they know little about—to set strategies. And yet, companies won’t experiment to find evidence of the right way forward.
Sadly, we the people are lazy and we’re after the one true answer. We want it now, without testing and experimentation. We want to believe our intuition or some experts’ opinion is the final truth. It’s not. Ariely again:
When we pay consultants, we get an answer from them and not a list of experiments to conduct. We tend to value answers over questions because answers allow us to take action, while questions mean that we need to keep thinking. Never mind that asking good questions and gathering evidence usually guides us to better answers.
This study shows that an incremental 241 percent ROI can be generated by applying data to business decisions. And 91 percent of CMOs believe that successful brands make data-driven decisions, as per Columbia Business School.
If your website was built based on intuition, HiPPO and mere best practices, you’re losing money every single day. You could be earning a ton more, but your website sucks.
Stop paying attention to your competitors and best practices, and start benchmarking against your customers’ expectations. Can you meet or exceed those expectations? Your conversion rate will be a leading indicator.
Build – Measure – Learn
Ever thought that Lean Startup methodology and conversion optimization are similar? You’re right, they are. In fact, I’d even say that the overlap is so huge, they’re almost the same thing. In the end it’s about growing a company by figuring out what the customer wants and how they want it. (By the way if you haven’t read The Lean Startup yet, pick it up now – it’s amazing).
The critical piece of the puzzle is this diagram:
This is the process to follow on your website too. You have some ideas (based on data – both qualitative and quantitative). You build a website – and measure *everything* that’s going on there. You do whatever it takes to gather useful data – web analytics, user behavior, usability tests, email tracking- everything. You end up with a bunch of data, learn from it – and form new hypotheses for new experiments. And repeat, repeat, repeat. You just keep experimenting.
Big data and research for research’s sake are stupid. What matters is what you do with the analysis and data you have – you don’t want to be “data rich and optimization poor”. One execution is worth more than all the unimplemented analysis in the world.
Structured approach to testing
Split testing via spaghetti method (testing random things and seeing what works) will only take you so far. Companies that employ a structured approach to testing get way better results.
Continuous optimization starts with a strategy – having an idea of what a successful conversion optimization system looks like, how to define the right goals, prioritizing where to test, how to plan for rapid-cycle testing and so on. Companies that have a structured approach to conversion are twice as likely to see a large increase in sales.
- Measure. You can only improve what you can measure, so measure everything. Be clear on your business goals, benchmark your competition for ideas, dig in your web analytics data, conduct customer surveys and analyze search behavior on your site.
- Analysis. Once you know your goals, it’s time to figure out what’s working well, what’s not and why. Analyze your content for relevancy and clarity, figure out if it matches user needs, do usability testing and analyze user paths/journeys on your site.
- Test. A/B and multivariate testing are the two most valuable methods for companies to improve conversion. Prioritize tests by potential value and cost.
- Optimize. After conducting tests, implement successful design and content changes.
It is a rather simple concept, but it is also complex as you can go very deep in analysis (too deep for your own good). To get the most out of your analytics — or just your optimization efforts — develop a cost-effective, smart system for improving continuously.
Conclusion
Your website budget needs to account for both – getting the initial version done and running continuous optimization programs on it for 12 months or more. That’s the way to keep growing your sales while minimizing waste and achieving your potential.
Optimization never ends since the world is constantly changing – user preferences, devices they’re on, their mindset, market conditions, and so on. What worked yesterday won’t necessarily work tomorrow. Keep experimenting.
Any chance of an online course or a blog post about exactly what do measure, how to measure, what tools to use and how to understand what data is relevant and what is not?
For example, my website. According to crazyegg scrollmap, my “add to cart” button is in the coldest area. I realize that I should do a/b testing but why would I show it before user understands my product value? Or should I? Or maybe the problem is that I have a onepager and people get tired of scrolling? Or this scrollmap data is not really that relevant at all?
That’s why a detailed post or screencast of what data you measure and based on that data what kind of decisions you make, would be nice :)
Thanks. Great post btw.
Margus, I had a look at your site and I see what you mean you have a good landing page but your main CTA is buried some visitors of your site will have already heard about you and may be looking to buy straight away while most others will be ready to buy at different points on your page so you need to have CTAs at strategic place all across your page but probably no more then 3-5
Yes – conversion course is in the making!
Anxiously looking forward to it :)
Looking forward as well! Just want to get to a point where I can hire you guys to do what you do best! Your content is enough to change anybody’s business….thanks so much and keep doing what your doing!
Agreed! Continuous optimization of a website is important to be able to be and stay competitive in your market. Design is just one aspect of a great website, SEO and content are equally important.
Great article and great reminder that optimization is an ongoing process – not an end goal. Thanks for sharing!
Really good article and I guess I’m as guilty as anyone else for not implementing continuous optimization. Actually when I think about it, it’s crazy not to focus on this. At the most basic level, surely it’s easier to increase conversions on existing traffic than it is to spend money on SEO and PPC or go chasing after the latest, “hottest” new free traffic source like Pinterest.
Add me the list of those looking forward to your conversion course.
Regards
Great article Peep. I have also found it amazing how so many accomplished product managers who employ A/B testing and use optimization software can stop short and decide to use guesswork for a given feature (i.e. my users all want to login with Facebook or Twitter!).
I also agree on the need for continuous optimization. My team and I have seen numerous customers caught off guard by an unintended or intended change to their product or audience and see conversion rates suddenly plummet for an “unexplainable” reason (we very often see this stemming from changes in traffic sources – new ad channels or keywords).
As such we advocate a programmatic approach to testing which uses MAB algos rather than repeated A/B tests to be able to continuously monitor and adapt content without having to create an ad hoc program of re-running static A/B tests on each variable.
Would love to hear your thoughts on this approach.
All the best!
OK, so we all know these principles. I’ve been in optimization at least for 6 years now. I live and breath this thing! That’s why I get so disappointed when companies get fooled by fancy design pitches by agencies. Yeah, put a banner on the home page that covers half of your site with no copy or CTA. Awesome(!) Looks fancy though!!
We need to bridge the gap between optimization principles and the design process. They are still very much parted and that’s one of the rootcauses of problem. I don’t see any design agency doing a realistic pitch saying “We are going to update your design and that’s probably going fix %40-60 of your site issues. Once the site settles down after redesign we are going to optimize the design with the data and customer feedback we have for the next 6 months. That will get us to %60-80 at best. The rest will be ongoing cycle of learning, optimizing and growing.”