Designing your sign up page

Turning a visitor into a user of your application is difficult. Turning a user into a customer is even harder. Much has been written about designing for sign-up, focussing on funnels, metrics, cost-per-acquisition, etc. I even wrote an article myself. A lot of the guides focus on the assumption that lost customers are a result of poor form design, bad layout choices, and visual design related blunders. Unfortunately that’s very rarely the case.

Sign Up Funnels: Myth and Reality

In some transactions you lose customers at each step in the process. In retail stores, this can be measured. Gap know that the more customers who try things on, the more they’ll sell, so they encourage their visitors into changing rooms. They know that by having an assistant on standby with a size up, a size down, and a different colour, they’re more likely to close the sale. They know how attract and aquire customers.

The only equivalent of this is shopping carts where you can measure conversions by look at customers moving from one step to the next. It’s a lovely idea that a purcahse works in a perfect measurable funnel “View Product -> Add to Cart -> Go to Checkout -> Enter Details -> Confirm Order -> Success”. That’s true for some cases, but most of the time there’s five tabs open looking at different prices/charges/delivery dates/ refund policies/ taxes etc. And then purchase might happen once they get home, or when they get paid, in a separate visit, possibly recoreded as a separate visitor. That said, shopping cart funnels still offer heaps of information that should inform design decisions.

The problem is that none of this works well when designing marketing sites for a web app about to launch. There are different forces at play.

The cheapest app money can buy

Web applications are rarely a commidity. Commodity web apps are things like file format conversions, URL shorteners, Twitter pic uploaders, File-hosting sites. They’re disposable one-off transactions and the user doesn’t really care what URL they get out of the exchange. They’re tough rackets to be in.

Users aren’t looking for the cheapest app, they want fastest, more reliable, best supported app. The only thing that matters when designing your sign up page content that supports their desire. Are you convincing users that your product does something useful for them. Does it make them rich, make them laugh, pique their interest or get them laid?

If you offer me an invoicing solution and print “easy to use” everywhere on your site, it means nothing to me. Just like everyone thinks they have a good sense of humour, everyone thinks their software looks good and is easy to use.

Good Content sells

Content is king on marketing pages, yet often they’re the most content scarce pages on the web. A fancy tilted screenshot and a big red button doesn’t convince me of anything, except your ability to rotate images. Here’s an incomplete brain dump of questions you need to answer if you’re selling invoicing…

Your app looks simple to use but is it powerful enough to handle my set-up? Does your software know about the taxes/rules about how invoices are handled in my country? Will I be your first serious customer, or do have experience dealing with firms of my size? Do other firms like mine use your software? How long have you been running? How do I know you won’t wrap things up in a few months? Can I trust you guys? Can I talk to you guys? How do I know you are legit? Do you offer good support?

Begging your visitors to take your free trial is often the wrong approach. A free trial costs time and doesn’t answer all the questions. Screencasts are good, but they’re usually not enough.

Bear in mind also that invoicing is a well defined problem. People know what to expect of invoicing software. It gets harder when you’re pitching a solution to an unknown problem, or re-defining an existing problem. Take FlowApp for example, Flow aims to change the way I work. This means Flow needs to convince me that they know how I work, convince me there is a problem with it, explain how they solve it, why it works, who it’s aimed at, and then go ahead and answer all the other questions I listed earlier. No wonder they’ve yet to launch a marketing site. This is hard stuff.

What you can include

Before you open “Ye Olde Web App” template and routinely drop in the obvious components, think about how you would sell this to someone. What sort of information pushes people over the line. If you were trying to impress me at a conference, what would you say? Easy to use? Heard that before. Convenient? I’d hope so. You need more than that to attract interest, here are some ideas…

• What interesting figures can you aggregate (100,000 hours billed, 2,000 companies managed, 3.6 Terrabytes of data secured)
• Who’s currently using it, and for what?
• Who is the team behind the application?
• How long has the application been worked on?
• What significant changes has the app been through while alive? What is the story behind the application?
• How can you be contacted? Can you be called? How good is your support?
• How secure is my information?

“But many popular web apps don’t do this!” you might say. Firstly well established web apps are feeding off their recommendations and the established reputation of their creators. When you’re just getting started things are different. You might not have an audience yet, so unlike the big names, you need to win trust and respect. You might get the benefit of the doubt, but you can’t rely on it.

Secondly, many of the big name web apps have content heavy homepages. Look at Highrise or Basecamp, Mailchimp, Campaign Monitor, they’re not scrimping on information. This goes beyond “Content is King” and isn’t really about design. It’s about the ability to sell. Even when your product is stunning and sells itself, you still need to sell me on your company, your support, your features, your future. That’s why it doesn’t surprise me to see companies like 37Signals continue to add content such as the Yes page, or the customer support happiness page. There will be more to come.

The Exceptional homepage

The Exceptional marketing site has gone through many revisions over the past three years. One lesson we’ve learned is that the more useful information we can give visitors, the better our conversion rate. The numbers back this up. Each piece of content there is to answer a question. Our wall of logos lets you know that we are for real, and have 6,000 people relying on us. Our status site lets you know take performance seriously. Our blog lets you see our customers and what they use us for. Our features page details every single thing the app can do for you. Our screenshots offer tooltips to explain what you’re looking at. The point being every piece of content is there to answer a question, and bring you one step closer to sign up. When we discuss the site, it’s from a “What else would persuade people to sign up, if they knew about it” Our last addition was the row of supported languages & frameworks, and again we’re seeing positive results. Allan Branch of Less Accounting reported that adding a phone number increased conversions by 1.8%, we’ll look at that next.

Metrics for marketing pages

Metrics are great for telling the what, but not the why. No matter how many Google Analytics tutorials you follow, you’re never going to find the killer regex that checks for “user actually being interested in the app“. Your best bet there is to start finding people who you know should be interested in your software, try to sell to them, and find out why works and what doesn’t. If you see a lot of inertia, lots of “I can’t be bothered” then you have two choices. Either target new consumers in the market (i.e. the people who have no solution at present) or identify a new feature that users will pay for. Be wary of the latter tactic though. The world is full of people would would buy it if….

As I’ve said before, the truth with funnels and A/B tests is that they’re of little value during the early days of a web app, when traffic isn’t significant. I’ve seen many A/B test junkies wait a long time for customer ‘B’ to even show up.

When you don’t have the volume, go for the personal approach. When you can no longer go personal, then analyse the volume. At every step you need to ask yourself “Is every single thing on my website selling the product” and “Is there anything else I can include that will help“.

Looking through Mixergy interviews with successful founders, you could be forgiven for thinking you needed a popular blog to be able to release an app successfully. The correlation here isn’t coincidental. A popular blog is an indication that the writers can sell things, whether it’s their credo, depth of thought, technical skills, or opinions about business. It’s surely no surprise that if they can sell themselves they can also sell their software.

The thing is we’re all salesmen, and whether we like it or not we’re always selling. We just don’t wear the shiny shoes.

Follow @destraynor on Twitter, or join the discussion at Hacker News.