Bubble Service Level Agreement: why doesn't it exist?

[georgecollier]

The last few months of Bubble development have been pretty awful with all of the editor issues, run-mode bugs, server issues.

@josh (and now @payam.azadi) have consistently said how important stability is to Bubble:

@payam.azadi

The approach we plan to take is one where we are consistently making gains in our service reliability. I’m blessed to have an incredibly talented team who shares this vision and is working tirelessly to achieve it.

@josh

In Q4, we plan to spend even more time on systematic improvements. We are planning to spend a substantial fraction of our Q4 roadmap on improving observability (which helps us catch performance regressions and new bugs), testing, and stability / robustness improvements, focusing both on the overall platform and on the Bubble editor.

But it’s all talk. We’re certainly not seeing any of that. It feels like any Bubble engineer can deploy a change without any testing.

None of these ‘plans’ have any substance without an actual SLA. Bubble needs to put its money where its mouth is and deliver an SLA that commits to a certain amount of uptime and a degree of stability in all aspects of the platform.

Right now, it’s just promises and plans which get broken time and time again.

If Bubble genuinely believes that it is competent enough to deliver good service is, then they should prove it. Otherwise, Bubble’s just showing they don’t believe in their ability to do it. If they don’t, why would we?

[S2294]

Its a well known fact that Bubble employs a large number of ‘interns’ who often release features which have proven to be nothing but ticking time bombs in terms of platform stability.

That is just greedy on Bubble’s part. Getting juniors to work directly on platform features affecting millions of users for next to nothing pay.

Even when Bubble announced they were jumping on the AI bandwagon, they publicly announced they were offering $30k for a 3 month AI study. Like come on, why not hire a subject expert full-time, in house, if they were serious about AI? (which should be the last thing they should be working on where there are tons of bugs they haven’t bothered fixing for years).

The truth is Bubble is a mess under the hood and offering any SLA is akin to shooting themselves in the foot. Since they took on $100m in funding in 2021 they have been doing bare minimum fixing bugs and pushing their interns to ship new features to get new customers. They spend more time advertising on Facebook than fixing their bugs.

My only hope is the new director of engineering who joined Bubble a month back actually fixing the technical mess under the hood, which will of course come with more bugs and downtime in the short to medium term.

[georgecollier]

@S2294

The truth is Bubble is a mess under the hood and offering any SLA is akin to shooting themselves in the foot.

My thoughts exactly and I don’t see how they can prove they’re reliable without implementing one

[artemzheg]

First of all, I’m sure Bubble has internal SLA for incidents and SLA for Enterprise plan.
Secondly, to introduce SLA for non-Enterprise they’ll need much more support engineers. Otherwise SLA will be always violated (3,600,000+ developers).

I hope someday SLA will be announced, but I’m not sure if we will see it in the nearest future.

[georgecollier]

@artemzheg

they’ll need much more support engineers. Otherwise SLA will be always violated (3,600,000+ developers)

It doesn’t even have to be that deep. I’m just looking for something that commits to 99.9% platform uptime and gives platform credit if it’s not met. Obviously for something like the editor it’s harder to quantify, but at least for the hosting…

[lizzie]

Just wish they had a proper system for engineers on call and a support team who have the balls to actually reach out to the person on call when a situation happens. Time and time again, these issues happen outside of EST timezone and time and time again we’re left waiting for New York to wake up. It simply isn’t difficult to put such processes in place.

There is either a lack of training for support agents to use the escalate and hit the wake someone up button or the processes are not setup correctly to identify critical situations.

[artemzheg]

@lizzie

Time and time again, these issues happen outside of EST timezone and time and time again we’re left waiting for New York to wake up.

Don’t remember where but not long ago Josh posted that they 24/7 have a duty engineer to check platform level issues and wake up the relevant developers if there is a need.

P.S. It’s here:

Page data not updating

Following up on some of the questions and concerns – Re: … that’s a great point – that was our main hesitation with this before, but we’d likely be comfortable releasing it on the new WU plans. I can’t promise we’ll prioritize it immediately, because given the way our client-side caching works, the implementation of this feature isn’t completely trivial, and we’re trying to carve out space in our roadmaps for more reliability, observability, and bug fixing. But I will highlight that feedback …

[chris.williamson1996]

At this point with how unstable it is on a near weekly basis they need to start allowing people to choose when ALL editor updates come like you can on dedicated instances.

Since we moved off of our dedicated instance bc it cost $13k/3mo I realized why they pushed the fact that dedicated isn’t effected by updates the same way.

There is no reason the public instance should be this unstable if they already have a fixed process and SOPs they use to avoid it in dedicated.

[georgecollier]

@chris.williamson1996

There is no reason the public instance should be this unstable if they already have a fixed process and SOPs they use to avoid it in dedicated

I’m not technically qualified enough to fairly judge but from a Bubble user’s perspective, it feels as if Bubble’s underlying engine (both run mode and the editor) is built on a house of cards that could all just come tumbling down one day. It doesn’t particularly matter if that’s true or not, but if that’s how it feels, you can bet I’ll avoid Bubble dependency in the future…

[S2294]

@lizzie

Just wish they had a proper system for engineers on call

@artemzheg

duty engineer to check platform level issues and wake up the relevant developers if there is a need

It just sounds ridiculous that anyone needs to be ‘woken up’ to fix something in a software company with millions of worldwide users, in 2023.

They should have teams working round the clock in overlapping timezones. Every other software/tech company has engineers working remotely from multiple parts of the world that covers many timezones. And its not a cost problem, its cheaper to have staff in developing countries like India, Latam etc… Its a cultural problem. Infatuation with holding users around the world to New York timings, shipping untested work of interns which always breaks something and treating production as a test environment is a cultural problem that has been played out over many years for us users to see.

@josh @emmanuel it would be great if you could address these issues in your monthly report rather than what % of coffeescript code Bubble has. Beyond ridiculous to be honest.

[artemzheg]

@S2294

They should have teams working round the clock in overlapping timezones.

That’s something we’ve been waiting for ages