Recently I've been coding some of my design prototypes in ReactJS to combat these issues. React is built with components rather than pages, and is stateful at its core. Of course it's not without its tradeoffs, there is a steep learning curve and some things can be difficult, even things that are trivial with other tools.

It's not for everyone, but if you're interested I wrote a short post talking about how and why I use React for design.

I'm not a data expert but I have some thoughts on this.

If you're looking to affect change rather than just report on what's happening, try to reduce down the raw numbers down to a more useful metric.

The raw numbers can be a vanity metric, or even worse, they can give a sense of overall success while masking problems such as:

• haemorrhaging users at a problematic step in the funnel

• a slowing rate of growth

• a decreasing overall completion rate – growth at the top of the funnel is faster than growth at the bottom (this can be a sign of marketing reaching a poor audience)

For example, if you want to increase the conversion between particular steps in the funnel, show the calculated conversion rate and its change over time (or a comparison to a previous period).

If the number of users completing the funnel is growing, and you want to achieve a "hockey stick"/exponential growth curve, plot the growth rate and change over time.

If the funnel is experienced over a longer timespan (eg product onboarding and activation), compare cohorts of users to help pinpoint which product changes impacted the conversion rates.

Most of all, avoid the traditional funnel-shaped visualisation. The weird shapes obscure a viewer's ability to understand the data at a glance. Stick to bar charts to show discrete steps, numbers to show important values, and line charts to show values changing over time. And remember to add goals or thresholds to give the numbers meaning.

Not a back-end dev but I always learn best by trying to build things.

In terms of language, Python and Ruby are both pretty safe choices I think but if you're a front-end dev already then why not Nodejs? Javascript everywhere.

Yeah Axure and other visual prototyping tools are great for quickly blocking out and iterating on the core layout and functionality, I just don't think it's the place for advanced prototyping with conditional logic, handling data, fancy scroll effects, drag-and-drop etc. Just because you can do it in Axure doesn't mean you should.

The only reason EU-based airline sites and the Australian Jetstar site are any better is because of regulation. They would be using all the same (and more) dark patterns if they could.

I've done some crazy stuff with Axure. It has the ability to do a lot more than other prototyping tools - it's programming, but with a UI. But if I was starting from scratch I would focus on learning to prototype with code instead. A framework like Tachyons coupled with some javascript (even just jQuery) will be much faster, more flexible and a more useful skill once you get the hang of it.

There is Origami for prototyping but I've not seen anything for production.

There seems to be a tendency to think that every app is unique and can't be made by stringing together some pre-configured (but customisable) building blocks in the same way a game is.

I like the idea though. Are you going to make one?

Good artists/great artists and all that...

Actually some pretty interesting typefaces. Others - https://www.behance.net/hankendesignco