@MKE_Howard wrote:
But we have to remember that the forum is changing for a reason.
What was that reason? I've yet to hear much more of a reason, than "we're updating it"
I don't see resistance to change being the primary factor here, and I think we'd be missing the bigger picture to view it as just that.
Minimum Viable Product
The mistake made was starting with a MVP ( minimum viable product) that was far too minimum. MVP is an outstanding approach, when done well. It allows us to stand up a new design or product, and begin getting feedback about it right away, rather than spending a lot of time/money designing features that users don't care about, or that the market doesn't support, etc.
Again, it's an outstanding approach... when done correctly.
If you're doing an A/B test where you are comparing the new (B ) to the old (A), you really should never start with an MVP for the new product that is less finished or robust than the existing product.
A/B done well
In order to get user buy-in, and a final product that focuses on improvement and optimizes the amount of improvement that comes with each change, you'd want to make sure the new product is equal to the existing product. Or better yet: it demonstrates some minimum improvement over the existing product right out of the box. You want users to be excited about the improvements and not fearful that the changes are going to result in a lesser product. And it should really be voluntary testing, not forced. If done well, people are generally excited to be part of the test group, so they can see the new stuff.
That's the way an A/B test should look: the new B product, when it is introduced to users at the beginning of the A/B test, is at the very least, equal to the old A product, and preferably it demonstrates some minimum improvement.
A/B done poorly
But imagine if we just went with a MVP that didn't consider the existing product. On it's own it works as a minimum viable product to get us up and running. But when compared to the existing product, it looks janky and unfinished. The userbase rejects it from the beginning, because we're providing them a far inferior product in comparison to the one they already know. There's nothing for them to be excited about, only things to be disappointed by. We've shot ourselves in the foot, right from the start. And if we stubbornly continue with that A/B test, the data and results collected are going to be skewed, and we'll need to force a result. A result that has no chance of being the best product, a result that won't reflect the wants or needs of users, and a result that will likely be costly to maintain and have a quick obsolete date.
So if we've done that, we've crippled our MVP from the start, and then captured invalid results from the A/B test. It's a gross misapplication of some very effective development strategies, that will result in a product that simply never had a chance to deliver on what we originally set out to do.
Inventor Release as an A/B test
That is what we have here. This forum A/B test is the equivalent of if the Inventor developers decided to ask us to beta test the 2026 version of Inventor that only had the capability of Inventor 2016. If they attempted that, Inventor users would naturally only focus on what was missing, and what a step back it is.
Maybe those Inventor users are just scared of change though?