An Update to Workload, Plus More Transparent Calculations

[artemzheg]

@gaimed

Also paying more for Bubble.io allows them to hire more engineers and continue growth.

From my POV things should be done vice versa:

1. Bubble dev team refactors code to make most commonly used and blamed features efficient in terms of WU consumption
2. When #1 is done (or at least significant improvements can be seen) → Bubble switches all new apps to new pricing model and gives a XX legacy plan period for existing apps.

Right now it’s a game of hopes. We are promised to see significant changes. But if we will not or changes will not be significant - that’s just a waste of time that could be spent on “migrating” apps to another platform (or going full code) but will be used to swim in the blue ocean of “optimising”.

[gaimed]

Well, they have a track record of being good for developers. They have enabled non coders to create great apps using no-code.

They have made some mistakes but overall I trust them… but everyone should decide for themself!

[artemzheg]

@gaimed

They have enabled non coders to create great apps using no-code.

That’s the point. The ease of making apps on Bubble and cheap pricing were a catalyst for the growth of the Bubble’s popularity. But with new pricing model and WU concept things become more expensive and unpredictable. Especially for newcomers and inexperienced devs.

[boston85719]

@boston85719

Bubble likely will not survive this anyway if they don’t show significant improvements within next 6 months. Can’t lose the thing (the community) that made your product special and stay the leader. Most listicle articles highlight Bubbles community as it’s core strength. They have already lost some pillars of the community because of this.

Just got a blast email from Bubble

Seems like they liked my quote

An Update to Workload, Plus More Transparent Calculations

I bet you the new feature of selecting which data fields to retrieve in search is rolled out in under a month. I’d expect to see everything from this thread in under 4 months. Bubbles got the engineers now, so a very different amount of available man hours than what they had in rolling out the flex box. In my 5+ years on Bubble, I haven’t seen as many new features as I have in the past 3 months.

Hopefully they won’t let us down @josh with significant improvements and a ton of features to help optimize apps with workload units in next 6 months.

[simon.wong]

I wonder how much of their new pricing structure has to do with the fact that they’re tied to AWS. The WU pricing seems to very much align with AWS’s pricing model where they do charge on volume of usage.

[Ren]

Hi @mac2, can you (or anyone else) help me clear this too?

here is a copy and paste of the questions

For the new plans, there is no longer an option to purchase Server Capacity Units.
This is because "they will all have the "highest capacity that Bubble can offer while maintaining the integrity of our own infrastructure."

In light of this I keep wondering:

• 1) How much is this “highest capacity…while maintaining the integrity”? I don’t think that an app can be left with all the hardware availability in the meantime unused by other apps residing on the same server virtualization, and for that matter, the reference itself to “maintaining the integrity of our own infrastructure” implies that the speed limit is also in place for future plans anyway.
  If I am not wrong (and I certainly hope I am) I’d need to know: will the new “highest capacity” be closer to the highest capacity today offered under the current Professional plan (10 Units), to that offered under the Production plan (40 Units), or will it be a new limit identified by Bubble (perhaps somewhere in between these two)?

• 2) It has been made clear that the new plans will no longer have the ability to purchase server capacity addons. Can we rely for this on the fact that those paying $32/month for the Starter plan will have the exact same hardware virtualization as those paying $134 for the Growth plan?

I have been trying for more than 10 days to get these answers, which in my case (but I may not be the only one) are essential to making a decision now about the next 18 months for which I have about 24 hours left.

Thanks in advance!
Renato

[gaimed]

They switched from a “server” model to a serverless model. So instead of paying for capacity you are paying for usage. Serverless can scale fast while servers are capped at the limit.

So there is no limit anymore or capacity or hardward virtualizations. It will be usage (calculated as WU’s).

[kurt_fireball]

This pricing model means they can increase the price whenever they like, without increasing the price. Just re-adjust the WU weighting. Nobody would ever know.

[Ren]

Thank you very much @gaimed !

@gaimed

They switched from a “server” model to a serverless model. So instead of paying for capacity you are paying for usage. Serverless can scale fast while servers are capped at the limit.

So there is no limit anymore or capacity or hardward virtualizations. It will be usage (calculated as WU’s).

Empirically I’d be inclined to think that “serverless” is just about “remodeling” server provisioning, usage, resale, etc., so to avoid keeping resources allocated for the hundreds of thousands of free apps left to vegetate by users in their accounts, generating costs for Bubble - and that frankly seems more than fair and sensible to me.

In any case, this is not my field at all so I am careful not to push these lame reflections any further. Therefore, returning to the ground: can we take the following statement as definitive?

The 32 USD/month of the Starter plan, the 134 USD/month of the Growth plan or the 399 USD/month of the Team plan offer exactly the same performance in terms of processing capability/speed , etc. - the differences between the available plans being related only to features and WUs included.

Or to put it in the automotive metaphor chosen by Bubble: we can definitely say that whatever plan we subscribe to, we will all have the same car, where the engine will be absolutely the same for everyone, but as you go up in plan you will be entitled to more optional extras such as leather interior, air conditioning etc. (i.e. app editors, custom branches etc.) and more prepaid fuel (i.e. number of WUs included)

Do @josh , @emmanuel or anyone from the @bubbleteam feel like confirming/correcting the above statement, with only a handful of hours to go before the April 30 deadline - or alternatively postpone the deadline until it is clear what is included in the new plans regarding this specific area?

[gaimed]

Yes. Because they already confirmed that. So on the new plan all the plans have equal speed (but other features like more editors) and some WU’s included (in your analogy it would be called gasoline). You can buy more of those WU’s on higher plans or with WU addons.

[Ren]

@gaimed

Yes. Because they already confirmed that

Hell I must have missed it! Could you be so kind as to linking me to where they confirmed the above?

And, BTW, thanks again @gaimed

[gaimed]

here: FAQ: Pricing and Workload - Bubble Docs

What is the relationship between workload and speed?

Workload is a measure of the work that Bubble does to power your application over the month. There is no direct relationship between workload and speed. However, apps that were previously throttled by capacity will experience significant improvements in speed after switching to the new pricing plans.

Also on other places its confirmed that there is no speed throttling on new plans anymore

[Ren]

Oh, I see. I had actually read those sentences, and also quoted them in my previous posts, but I personally interpret them differently.

In any case - taking into account that Bubble’s founders said they were following the conversation - I have to assume that their silence on the matter counts as assent:

“Sold at 32USD/month, the Starter plan offers the exact same level of performance (i.e. speed and processing capacity) as the more advanced plans - Growth, Team or Custom.”

Amen.

[gaimed]

Its a bit to much to expect them to respond to so many messages. You can email support with this question if you like triple confirmatiom

[Ren]

Hi @gaimed , just to leave no room for doubt:

@gaimed

You can email support with this question if you like triple confirmatiom

I obvously did it beforehand and, since I didn’t get any answer, I asked it in the forum as well.

@gaimed

Its a bit to much to expect them to respond to so many messages.

I would say that this is not my company nor my job, however I read them. We have different views on this, but I would say it’s just fine.

In any case, April 30 has passed, and we have all made our own choices about the changes introduced, based on how they were given out.

Thanks again for your contribution!
Renato

[Bubbleboy]

to the point highlighted by @Ren “Sold at 32USD/month, the Starter plan offers the exact same level of performance (i.e. speed and processing capacity) as the more advanced plans - Growth, Team or Custom. ”

Has anyone upgraded from the Legacy Plan Personal to the New Starter plan and seen and improvement. I am on the Legacy Plan Personal and don’t see the ability to Boost.

If this was mentioned elsewhere… i’ve not seen it…

[jgellis]

Comparative App Cost – Old vs New…

As I mentioned in an earlier post we are launching or new app shortly and thought the community would be interested to see the impact of WU metrics. We just completed the wrapper (Thank you BDK), and are rolling it out to 100 Alpha testers. The following data is with our first three testers.

As you can see from Figure 1. & Figure 2. our projected server loading was a approx 5% (peak at 5.7%) with our three users constantly loading the app. (IE: Clicking constantly to force max DB read/writes and backend server workflows) Based on these figures our estimated concurrent user load max loading would be about 200 – assuming normal user click rates (Eg X/4 with a degree of error to favor performance). This means that we should be able to support 1000-2000 clients assuming only 5%-10% are active at any point in time. BUT NOT SO FAST. Now take a look at Figure 3.

You will see that over the short test period May ( 2, 3, 4) our total workload usage was 27,346 (approx: 3K WU per day per user). This means if we had 200 normal users they would likely consume almost 4.6M WU per month (3,038/4 * 200 * 30 : Average Units/normal factor times concurrent users times 30 days in an short month). This means our projected cost would go from $1,548/yr to $7,896/yr - a 410% cost increase! These data is not “make believe” they are factual results reported by Bubble.

Another way to look at this is how can less than approx: 1.45% average daily server usage (integral of hourly usage/hours) translate into 3.65% allowable WU consumption (9113 / 250,000) of the Growth Plan (comparing that to our current Pro Plan). Either way you look at it, you do not have to be a rocket scientist to understand that “Something is rotten in the state of Denmark”.

John

PS: Before anyone suggest our app is not optimized - please don’t. I can assure you my team has significant experience and does understand best practice. That said, Bubble does not offer all the tools a programmer would leverage to optimize code flow such as simple If Then Else, and Switch constructs. Instead they opted for When statements tied to every element that must test the condition at each step. Yes, you can use trigger events that act like sub-routine calls but this does not negate the downstream conditional tests that do add to WU counts… Nor does the lack of the ability to proactively index data tables or manage joins, to optimize DB performance help… but the lack of conditional, other common flow constructs, and DB tools is a topic for another day.

Figure 1. Maximum Capacity Chart

Figure 2. Percent of Max.

Figure 3. New Work load chart.

[Guillaume.b]

Have you got any answer from Bubble? Have you tried contacting them? If you do, can you please let us know what they say?

[davidthorpie]

@josh I had intended to scale my business on bubble, but looking at even the latest WUs for my app, which is still being developed by ONLY me - by the time I have even 5 users, I will be looking at hundreds per month in bubble costs - before having a single paying customer! This is not viable, not reasonable, and impossible for any entrepreneur to absorb without investors.

Having invested countless hours, over many MANY months, and only thanks to @keith’s incredible plugins and “lessons” (via forum posts and videos) that have helped me workaround the countless limitations in bubble’s client-side capabilities, I’ve been able to optimise the app as much as possible. I review the latest workload graphs, and not a single query can I optimise any further without dropping key features or dramatically punishing performance due to the data layer abstraction. Yet still, the app is unviable on this platform with the current price structure as it is.

I built my app on bubble with the help of an agency (initially), who swore by it - and speaking with them recently, even they have moved away from bubble for their new clients. As a full-stack platform, the enticement of bubble was fantastic and achieved my goal of launching without the need of native code and devs - my app has been heavily built by ME, a guy with far too many things on my plate who hates writing code, without a team of devs. It is incredibly frustrating and sad to see the way this structure of pricing essentially destroying my hope to initially scale on bubble to a few hundred customers. If I did this now, the pricing will sink the business before it even begins.

Thanks to this pricing decision, my entire launch plan is now on ice - as I begin the arduous process of researching the market, and having to rebuild the entire app in either one or more competitor platforms. Couple that with the snail-pace client capability innovation, the low-quality native plugins, the lack of regional servers, lack of basic CSS and scripting capabilities, minimal native capabilities, etc. - the list goes on - it’s just not worth sticking around at this price. If there’s a massive price hike, at least show some genuine innovation to justify it! Not to mention the total absence of external code repository support or branches in lower tier plans. It’s just a slap in the face.

What Bubble really needs to understand - is the nature of their customers. It is true you cannot cater for everyone with a one-size-fits-all approach, but no-code is ultimately there for entrepreneurs out there, who have an idea but don’t have the finances or the know-how to do native code. Bubble was providing a genuine service that was game-changing for anybody with an idea - it was the MS Word for writing web apps. But this pricing structure screams “we’re all about shareholders and maximising ROI”, without any due consideration that it is customers like me or anyone else in this thread (or previous “frozen” ones) who are the ones that even brought Bubble the fame it has today. Upset the customers and price them out of the market - you’re left with an echo chamber and the “expected” ROI from all the VCs pinning their hopes on 10% returns will be for nothing. Not trying to school anyone, this is just simple fact.

I for one have had enough of the games, the excuses, the “justification”. Bubble has every right to charge more, as it was (that word again!) a genuinely a good service. I might continue with a dramatically cutdown presence on bubble for my future app for some components (probably identity only, since it’s pretty simple - but performance is pretty poor in my region anyway, thanks to a lack of local servers) - but ultimately, I don’t care for the added complexity of multiple platforms, and other platforms offer better client-side capabilities and native apps that could dramatically change the game for my company anyway - at a mere fraction of the cost of continuing with Bubble. It’s all a matter of time.

I’m frustrated, bitterly disappointed, and now also having to rewrite my business plan with the assumption of a non-bubble backend provider. No doubt, I am not alone in this situation.

[brookelustig]

I think Bubble must have decided to change who their customers are too large to medium sized businesses? Only logical conclusion with this extreme change in pricing. At least everyone has 18 months to find different solutions.