Why only small wins of late?
Perhaps this is a bit of my own selection bias, but it seems the majority of the new release announcements concern either back-end API improvements (which do not appear to lead to visible new features) or minor UX changes (yes, I appreciate that the screen scrolls x amount less on action y, but I probably did not notice in the first place). Dark mode has been the biggest item I have seen released (and yes, that was a win).
My surmise is that a heroic amount of work goes into Direct Import (which I do not use). Exposing the API seems like a cool win, but how many folks are actually registering and integrating with a RESTful API and bothering to create their own UI given that the Toolkit exists?
There are numerous features requested on these fora that never seem to see the light of day. How about a few longstanding pain points (I am sure others have their own):
* Inability to reconcile in the mobile or tablet apps
* Woefully constraining set of recurring transaction options
* Stealing from the future pitfalls that snag even advanced users
* Requiring intellectual hoops to deal with a credit card refunds or deposits (now far more common with rewards cards trying to find ways to keep travel-loving customers)
* Constraining users to budget month-to-month regardless of their pay cycle
* Reporting limitations so pronounced that an export to Excel is required to extract useful data
* Not allowing the option to record a credit card payment from the cleared rather than the total balance (only Amex allows me to overpay, which is a rare event)
* Lack of another dimension apart from category (tags or any other way to categorize business from personal from vacation for example while preserving dining, transportation and entertainment as bone fide categories)
* Not carrying negative balances forward (or at least offering the option to do so)
Minor UX features are great. Squash any and all bugs. However, my guess is that Direct Import consumes development effort at the expense of base functionality. For a simple proof of the need, look how many are using the Toolkit, which simply screen-scrapes the data already delivered by YNAB into a more usable form. It should be even easier for those owning the data to deliver the same and better features.
(Departing sofa and soap box; still love YNAB and hence the passion for making it more useful in such simple ways).