Ask HN: Can bad onboarding destroy user activation?

1 pointsby curiously11 tahun yang lalu6 comments

So I'm bootstrapping a SaaS that lets users upload videos and do data mining on it. I had a few customers so I know that it is working and providing value.

But what's happening next is really stressful. It goes something like this in a typical week.

I see a bunch of trial sign ups. I get excited.

User uploads videos but they did not configure the settings correctly so they end up with bad data.

User tries a few more times or gives up or reaches out to me.

The majority of people are just giving up and never returning.

The ones that try it a few more times ends up requiring extra features I need to build that will take a long time and they won't pay up front for it.

The ones that reach out, I teach them how to use it properly. Only a few of these actually become customer.

My question is, is my onboarding broken and is it stopping user's from becoming activated? Is the first experience enough to destroy the product's value proposition?

What about for the people that ask for features so they can get what they want out of it but unwilling to commit up front for it and expect me to work on it for weeks and still trust them to pay for it after?

OR is this a completely different issue? am I simply getting non serious people signing up and draining my resources and causing stress?

I do have some plans to address both:

1) hold webinars or do screen share demo as a mandatory step after signing up

2) get rid of free trial

3) identify new target customers and discover market channels to reach them.

What do you think? Really could use some advice here! Thank you.

1 comments

nostrademons11 tahun yang lalu

Speaking from experience trying out various SaaS products - if I hit an obvious bug or get corrupted data when I try out a demo, I immediately give up on the product and try something else. Reasons:

1.) It speaks really poorly of the level of polish that I should expect in the product later. You'd think that someone would at least run through the demo to make sure it works without error.

2.) Usually if I'm evaluating a product I'm evaluating multiple products, and also have the option of building my own on the table if none of them work out. An early error is just a signal to focus my efforts on one of the other products on the marketplace.

I think you might also have a problem with the product concept, if the folks who do try it without error want additional features or aren't using it even after training. But to answer your question - yes, a bad demo or onboarding experience will absolutely destroy interest.

curiously11 tahun yang lalu

1) That's what I suspected. 2) What if I force them to hold my hand so that I can show them to how to properly use it and convince them that it works?

Since first impression is so important, and the product needs hand holding to demonstrate it's value, my free trials would essentially be hurting me.

There are some video files that give a lot of trouble so it's really to support their very custom requirements as the product for the paying users works fine. It's those people that want the custom work done for free which I don't think is a great people to target.

xkcd-sucks11 tahun yang lalu

For users that abandon it, you could process the video that they've already uploaded and email the results to them, along with a short screencast/tutorial of how you did it? As long as there aren't legal/ethical issues with doing that.

Also, datamining is like 95% trivial and 5% picking the right parameters, so you might focus a bit more on making the automagic parameter selection a little more useful

curiously11 tahun yang lalu

yes I have done that before, record a screencast of how I did it and even email them the successful result. most of the time I got back no reply.

It works for majority of cases but still the tail scenarios where it will choke. those edge cases consume majority of my development resources. I'm at a point where I think I will just say, 'not yet supported but will let you know' and instead focus my efforts on the cases that I can actually support.

nostrademons11 tahun yang lalu

It can be useful to do things manually and hold users' hands at the beginning, so that you get a much higher-bandwidth feedback channel. However, if you want to end up with a viable product, the product will eventually need to be simple enough that someone can use it without you explaining it. Otherwise you have a consulting/systems-integration business, not a product business.

curiously11 tahun yang lalu

I think thats a good point. Eventually the product would've have to be a low touch, self service but for the niche I'm dealing with that's unlikely to be achieved with the resources that I have. Even when I do, I think there will still be those that fall through the cracks, I think that there will need to be a dedicated customer success manager or something like that to ensure that they are using it successfully and most of the time it's that they are using the wrong configurations or the source of the video file is extremely bad (choppy audio, weird artifacts etc).

The tail, edge cases make up a lot of the use cases (as I can observe so far), it almost always require additional development effort. However, I don't think I can do those work for free and hope that they will buy when it's ready.

I think I will actually consider removing the pricing pages, demo section and setup a "Call For Demo or Pricing" form that asks for phone number and company size.

I am focused on sales so maybe having a consulting + software type of business might work better.