[bacheson1293]

I have a $100M SaaS in a different space but the underlying tech is identical. I had this idea ~10 yrs ago when we first started writing our own import logic to move users over from competing services.

The engineering challenges to make this work flawlessly is going to be substantial to say the least. I GET from ~70 data sources right now...I couldn't imagine having to GET,POST, CHECKPOINT and DIFF with all the edge cases that appear.

Best of luck but this is a very hard problem to solve. Would love to see someone see this through!

[mbesto]

It's not even the technical challenge that is the big issue. It's the political one. Trying to get vendors to provide API access to migrate data away on a scalable solution will be a no-go for most of them.

[kevlened]

Co-founder here. We skip the politics because we're willing to scrape to export or import if APIs aren't available on either end. That allows us to function independently of vendor buy-in.

[buro9]

This assumes the incentive is balanced for the vendor, that they'll win as much business (or more) than they'll lose.

I don't believe this is true anywhere, but specifically those vendors who are most aggressive about creating walled gardens and who have high revenues with unhappy customers looking to leave are both the sweetest spot for your end user demand and vendor resistance.

I suspect that you hope to leave those vendors (when discovered) until later, once you have momentum. But I also suspect that these will be the ones end users demand very early on and that end user perception of usefulness of your service is linked to the difficult vendors.

Additionally you have a time and opportunity window problem. Changing vendors isn't core to the business of most companies, and instead it's usually the product of a concern at a point in time (i.e. costs this year) that comes with limited window in which to execute on the project (i.e. we'll only distract ourselves from our core business for the next 6 months to try and reduce costs, then it's back to core business). This is going to make end user acquisition harder (more expensive) for you than most startups as you're going to have to invest more in the marketing to ensure you're on their radar at that point in time (with the assumption you can execute within the window of opportunity).

I'm not down on your business, like others in this thread this is a thing I've touched at the peripherary of my career and I'm just sharing my thoughts. You've picked a hard problem and if solved it definitely has value.

[SamuelAdams]

Honestly if I were OP I would specialize in a certain industry, ie healthcare, manufacturing, apparel, auto, or similar. Then get familiar with the big software systems all these people use. Systems like Epic and ERP systems in general take companies hundreds of millions of dollars to migrate from and to. If their company can automate even a little bit of this, that would be very lucrative.

[hanniabu]

So if there's no API then you can't help them unless their data is public?

[shrimpx]

Presumably if a user’s data is visible in any way while logged in as that user, they can scrape it.

[kevlened]

That’s correct

[lucas_membrane]

I was in such a business decades back as a consultant, migrating clients from incredibly expensive mainframe systems to much more affordable alternatives, and I found out the path is fraught with perils. I was an independent consultant doing this sort of thing to get companies out of software contracts that would be around $10 million per annum in today's bucks. The vendor who was losing the clients had revenue approaching $billion/annum in today's bucks, ie a few thousands of times my revenue. I got a letter from their lawyers, an ultra-aggressive firm that had been written up in the NY Times for being so efficient at winning ultra-aggressive lawsuits. I took it to my lawyer, the best I could afford, and he negotiated a total surrender, which meant perhaps half of my potential clients were now off-limits. I did pretty well with the other half for around 10 more years, but the vendors whom I was helping by guiding their new customers through the inexplicable aspects of the replacement software, came to want my revenue and to prefer that their customers not be too well informed about what they were getting into, so they started putting terms into their software licenses that required new customers to fence me out. Game over.

I went back to work as a developer for a vendor earning huge profits in a different vertical market which they dominated. Their lawyers were far more strategically important than their development team. The amount of revenue they derived from capturing customers was audaciously incredible for a firm operating in so-called free-market economies, even though they were always up to their elbows in legal battles and skirmishes on multiple fronts over who owned what and who was allowed to use which licensed API's for what functions.

[MathYouF]

I feel like there's a deep learning (possibly computer vision) route to this that would be extremely hard to create but would be able to arbitrage 90% of the difficulty with updating for new UIs or handling edge cases. Had you ever pursued that route or seen things that tried?

[weird-eye-issue]

I really don't understand your usage of the word arbitrage here. Also computer vision is related to images specifically - how is that even relevant here?

[ManBlanket]

This guy can speak for himself but I've read the comment a few times and I would surmise there's some ESL word-soup happening here. I am going to hazard what he's getting at is using CV to import data from one UI possibly into another. I think to get around the lack of a lack of API support. Similar to the effect a company might use OCR software to migrate data from paper forms a CRM, or something? I'm still wondering why I honed in on this one comment, but hey you felt intrigued enough to leave a comment too, so there must be something to it.

[666b20753a29]

What does ESL mean? I can't seem to infer it

[nix0n]

English as a Second Language

[KuzMenachem]

English as a second language I presume

[Corrado]

English as a Second Language