test.ical.ly | getting the web by the balls

May/11

6

In a relaunch project: resist the urge to migrate every feature, instead lose some!

If you work on long running projects that live over a few years you will eventually be faced with the term relaunch. Some times there are technical reasons, most of the time there should be strategic reasons.

However the current system is usually feature rich and was shaped over a few years into its current state. One of the things you will hear from product managers and developers is:

We have to rebuild every single feature so the users are not disappointed after the relaunch!

Wrong.

Additional to the design costs, the possible costs for the new software and or agency helping you doing it you will always have migration costs. Taking care about content migration and eventually feature migration of things that can not be realised easily with the new system as they are applications on their own.

Only this migration cost can make up huge figures of money that will definitely not be welcomed when you have to show the price tag. They can also be extremely complex and put the whole relaunch at jeopardy of not being finished ever.

So what to do instead?

Think of a house move. You lived in the old house for years and you now want to move into a new one. Do you go into the cellar and pack everything to be moved into the new cellar? Leaving you with exactly no space just like you had before?

No! You clear out. You clear out everything you brought down there and never once got up again to use it. Everything you once thought you needed but never did can go in the bin.

Same applies to relaunches. Only that you need to find out what is not needed anymore.

Before you do the relaunch what you should do is prepare a big list of all features and then look at your figures and see which of these are frequently used by many visitors. If you don’t have the figures start tracking!

Some things will not be possible to measure like boxes appearing on certain pages that a supplements to a page. Here you should use A/B tests and see if removing it for some of your users you can notice a dent in your figures.

This focusses on popularity only. You should also try to measure what the monetary income of each of these figures is. How much ad impressions do they produce? How many subscriptions to newsletters, etc.

And how many costs do they frequently produce and will they produce when having to be migrated?

From all these informations you can then priorize each feature so you have an ordered list. Whoever does the business decisions in your company can now decide how many features from the the top of the list have to be migrated.

The remaining can be lost!

· · ·



<<

>>

Theme Design by devolux.nh2.me