My proposal last week to produce a web app instead of an iOS app met with a resounding lack of enthusiasm, judging from the poll I ran asking people if it was a good idea. Folks, let me do a little more convincing. Here’s a few more reasons for producing a web app:
- Google Analytics – you can list your website with Google for a service known as Google Analytics, which gives you a wealth of data about your users, including the browser used, the location of the user, the type of computer, whether mobile or desktop, how long they stay, whether the user is a unique visitor or a returnee, etc. That’s very useful data for targeting your users.
- A/B testing – this is a feature that is not possible in iOS due to the nature of the App Store. A/B testing refers to the ability to simultaneously run more than one version of your web page at a time. Why would you want to do that? Well, suppose you can’t decide between two versions of a web page sales pitch, what do you do? You let the market decide which is better – you run both of them at once, with people directed randomly to one or the other. Web pages are relatively easy to change, so as soon as you know which is the more effective version, you can make the change. Depending on the change, a page could be monitored on a daily or even hourly basis.
- Rapid deployment of app features and changes – each iOS app must go through the App Store approval process, which can take weeks, even assuming that Apple approves it right away, whereas rolling out changes for all users on a website is instantaneous.
- Apple doesn’t get 30% of your income – 30% is the going rate for apps that use the Apple in-app purchase program.
- Nothing for the user to download – removes one possible barrier to adoption.
- Can do SEO – it can be hard to get noticed in the Apple App Store, but SEO (Search Engine Optimization) can be done to enhance the chances of a web app being found. In addition, it is possible to test the effectiveness of your SEO using Google Webmaster tools.
- Google searches are more immediate – searches bring users directly to the app, rather than to an App Store link or company web page.
- Connect directly to social media on the web – get Facebook likes, Pinterest pins, etc., all of which are useful for marketing.
This is not to say that there are no challenges to a web app. Not all browsers render web pages the same way, which has to be accounted for during coding. However, Google Analytics will show what browsers people are using, enabling you to adjust for the major ones. Another concern is that older browsers, such as Internet Explorer 6, won’t work at all with the more sophisticated features of HTML5. In addition, not all desktop tasks can be accomplished using HTML. Finally, the device must have internet access to do any useful work.
Still not convinced? That’s OK, a little skepticism is in order. There are a lot of questions yet to be answered. On the balance, however, I believe that it is worth trying for the web version of my app.
You had me at Google
I find some of your arguments for a web app less convincing, e.g. time lost going through the Apple approval process. Are we talking Carbon 14 dating numbers here? If I am a consumer assigned randomly to A or B in A/B testing how will I know if there is a more effective version? Won't I just go away from the experience exclaiming this app rocks or alternatively, this app is clunky – forget it? I won't know the alternatives. I believe Donald Rumsfeld said it best with this mind scratcher:
"There are known knowns; there are things we know that we know.
There are known unknowns; that is to say there are things that, we now know we don't know.
But there are also unknown unknowns – there are things we do not know we don't know."
You're right, some users will get the lesser version, but the danger of not doing it is this: if you don't try, it may work out that ALL your users get a lesser version. Until you test it, you have no way of knowing. Note also, that the difference between the pages need not be dramatic. It may just be a matter of testing simple things, like testing the phrase “Buy Here” versus “Check Out Here”, or something like that. What we'll be testing for is not so much functionality of the website, but effectiveness in getting people to buy. That's something that can be measured with a high degree of accuracy.