This is basically a story of well connected people who used their connections and existing wealth to buy their way into a market. I truly despise this “method” of building a business, and I doubt it will lead to long term results.
These guys apparently had spent $160 million and according to the article, had “no product here.” It’s perplexing to me that they were then able to so easily raise hundreds of millions of dollars to buy competitors that did have a successful product.
And this was barely a year ago. Why do investors believe two failed founders will be able to grow and scale this business when they couldn’t grow and scale their own business?
I’ll personally make sure to avoid this company as I don’t at all respect their method for creating a viable “business.”
> Pulling off acquisitions at this scale takes a way more than you are giving it credit for.
I don't think anyone is denying that skill is involved, but skill is often orthogonal to utility. Skill can be abused. Con men, burglars, and pickpockets all have skill. Malware authors have skill. Torturers have skill. I'm not saying sales is equivalent to any of those (OK maybe con men) but the point should still be crystal clear. What people are saying is that Fivetran hasn't succeeded in creating any value, and might even have destroyed value. When a big company "succeeds" by acquiring competitors they quite rightly get antitrust scrutiny because that's bad for competition and innovation. The principle doesn't really change for smaller predators.
I think what they were saying is this company puts off snake oil vibes. It's like calling Microsoft a success in the game industry when they bought companies or products to position themselves as a leader of the industry when the argument can be made that they couldn't build it back up if it went under.
You seem to be focused on software development as the only 'legitimate' avenue to building a company/business; if this is true, why? Are VCs 'illegitimate' for being unable to run the companies and develop the corresponding products they profit from? Are software developers 'snake oil' vibe-y for being unable to create the hardware they rely on?
As a fivetran customer who did a bake-off last year between all the ELT vendors, there is absolutely “a product there”. The advisors’ comment assumes a lot of context which the article doesn’t 100% clear up.
Sure. Is that why Geoff Ralston referred to them as a cockroach, and surmised that everyone else thought they were dead until they pulled the rabbit (acquisition) out of the hat?
They took $160 million and turned it into nothing. Then they took $565 million and bought a successful company.
Summarized a different way they took $725 million and turned it into a business with $189 million in annual revenue over a period of a decade.
If they had just invested all the VC money they would have $1.8 billion. I’d consider that a more successful outcome to be perfectly honest. Is anyone really going to pay $5b for Fivetran at 26x revenue? Doubt.
Valuations are generally a multiplier of yearly revenue, which makes sense. Valuations are, on paper, what you'd buy a company for, and in the old school look of things, how much the company made per year was a good metric, and you'd multiply it by several years, since you'd probably not buy it as a super short term investment.
Now, the actual multiplier values used for that... we can debate them for ages :-)
Old school valuations are based on NPV discounts of future cash flows. That obviously includes profit, but also all the other factors around growth and drag along. You'd never get any of the valuations we see today based only on x'ing profit, even for the oldest of old school businesses.
> I truly despise this “method” of building a business
I used to think this way and then I realized "whats the difference between a company that raises $100M to put it all in sales & marketing, and one that raises $100M to acquire a business with paying customers"
The answer is the latter one is much easier to do.
At some point someone is purchasing the asset for $1B. If both instances get purchased for $1B, then why do you care how they got there (assuming they both did it legally)?
You're being tricked into thinking there is "one way to build a business". There isn't.
It's not that there is only some moral goodness in building a business from scratch. I hate the idea of rich people just using their wealth and connections to succeed, basically using their connections to get undeserved access to capital. Society would be better off if it didn't happen.
What's not especially useful in my opinion is wealthy people who are failing in their company calling up their buddies to get more funding to buy up other companies that are doing something useful and successful when your own product is failing. Regardless of this case, we are all better served - & society is better served too - by new companies creating technologies and solving problems that succeed or fail on their own merits; the group that is not better served by this is people who succeed through influence peddling and connections to other rich people.
I think that article misrepresents it a little, makes it sound like they had no product. Both the current and previous startup I worked at has been a paying customer of theres, at scale, making it about 6 years total
> It’s perplexing to me that they were then able to so easily raise hundreds of millions of dollars to buy competitors
Building things is definitely hard work! It sure is a lot easier to just know the right people, but you also should have the right skills and insight to identify the opportunities.
i mean, that was obviously hyperbolic framing by a forbes writer who needed to create some contrast in the before/after comparison in order to make the deal seem like a company-saving thing. company was probably already worth $4b even without HVR. they just exaggerated the deal difference based on available data points. dont take the article's hyperbole at face value beyond the cited facts.
Having a very short window of opportunity helps. In this case - they (investors) had only days to decide. There is just so much due diligence that can be done over the weekend. And the target was already making good money.
They didn't buy a competitor, they bought a company that was operating in a different market segment so they could cover both mid-market and enterprise customers. Neither product worked for the other segment.
I have a weird vibe with Fivetran, Snowflake and Looker together as a combo with their credit based billing. Those three tools provide a platform that makes it easy to forget about engineering things correctly and just throwing money at the problem. It’s makes it so easy to do really dumb expensive things. LookML is another “lock in” type product (much prefer using DBT with a Data Vault model). Once you’re knee deep in LookML you’ll have a hard time switching to another analytics tool. Snowflake makes it so easy to do massive compute that your badly designed data models won’t matter. Fivetran is really cool though, but you need to be careful about getting into bed with all 3 those products at once.
Who else do you use? Fivetran's syncing has been very reliable (compared to others we tried) but the "active row" based pricing is unpredictable and can have huge spikes in billing costs. I've been waiting for reliable competition with compute- or size-based pricing.
As a product suite it makes a ton of sense. Sucks for the downstream users needing to manage terrible data models and hacky ETL workflows in LookML though
Be careful with fivetran - they inflate your data and then bill you for it. In our case, they inflated by ~10x the record count by doing a join between two tables/endpoints which wasn’t required for data integrity. (I.e. there was a “balloons” endpoint and a “parties” endpoint and fivetran synced a non-existing “party_balloons” table.)
We dumped them once our first credit purchase ran empty.
We have been with them for 2+ years and are moving away from them. Great product but absolute nightmare on pricing transparency. Every year the contract is a fight with their bait-and-switch. Every year they try and double our price saying they changed their pricing model and we have not increased usage. Sales people are horrible and don't trust them at all. If you are a small company with a tight budget I'd stay away.
I can confirm I've encountered this as well. In our case it was with shopify. There are some product_* tables that cannot be turned off without turning off product entirely and they can be incredibly expensive to sync. It's a very poor experience compared to something like airbyte where you pretty much get full control and can even edit the connectors if you have the dev manpower. The fact that adding new tables to your schema is on by default also really rubs me the wrong way. If you don't disable this or very closely monitor what new tables get added a reasonable connector can suddenly balloon in costs.
Stitch can do a lot of the simple replication stuff for cheaper. Fivetran, apparently, is going to change their bonkers billing stuff (only 2-3 years old iirc) because they are losing customers over it.
i meaannn… i work at airbyte haha. happy to do a friendly chat if it’d be useful to you (im curious what motivates you to do this startup in a pretty crowded field)
Somewhere buried in the article is Airbyte's valuation. For a 1 million in revenue they had a billion dollar valuation. That's like 1000x multiple. Is it because they are open source?
Dumb valuation aside, Airbyte seems like it will wipe the floor with Fivetran and other pay-per-record data pipelining tools. Open source, decent community, you can host it yourself... Orders of magnitude cheaper.
I think it's because there are investors desperate to get into this industry but can't invest in 5T et al, so they invest in whatever they are able to hoping they'll capture some of the uplift
All of these visual ETL tools are nothing more yhan a patch to the new breed of analytical databases because their support for loading files sucks so badly.
Even the praised dbt is just a hack to make declarative incremental because the SQL capabilities of these databases is also bad and you need to template out your way in.
There will once be a SQL database that will make these things not suck and collapse this entire hot space of startups for data loading and missing features in Snowflake.
Snowflake will go away too if they keep ignoring so many "missing features" startups and not improving their loading or language capabilities. Unless they intend to spin up an ecosystem of "hacks" to complement Snowflake to make it usable, which is another reason for some saner system to displace them.
Disclaimer up front: I work for Snowflake but do not represent them here - I'm not in Product and have very little influence over direction.
What language and loading capabilities would you like to see us add?
I personally doubt we'll ever aim for postgres levels of extensibility, but Java has been in for a while and Python just went public preview - with great reviews from everyone I've spoken to.
Plenty of activity on the loading side too, although we don't really aim for visual ETL - you're pretty much correct that we're happy to let 1000 flowers bloom in that space.
Unrelated to the article itself but Fivetran has the worst possible product from a developer perspective, their APIs are unreliable and will fail on you randomly.
I understand the dynamics of articles in Forbes (the successful entrepreneur making his way to success, lots of them are paid for.) but most such companies have such shit infrastructure and APIs, I wonder if people are only paying because they're locked in or their non-tech clients use this stuff.
It seems every day I come across a new billion dollar tech company I've never heard of before. Their sites always have a ton of well known companies under their "Trusted By These companies" banner too, and always seem to include NASA.
Huge enterprises are often made up of thousands of sub-domains with relative freedom to choose different technology.
It's actually useful to remember if you're a start-up chasing a contract with an enterprise. You might bend over backward accommodating BigCo to get their name on your books thinking you'll make it huge once "they" start using your technology only to find it never actually gets further than 2 people in Sub-Department-Offshoot, and while you might get to use that fancy logo on your marketing material, you'll wonder if the concessions and vastly undercharged overwork was really worth it because the payday of actually having BigCo use your tech throughout their stack never materialised (because it never really does, but it'll be endlessly dangled in front of you to get you to agree to put in a bunch of features for free).
In my experience, there's actually gold to be dug from that relationship. The people who spend money in these BigCo don't spend their own money and their motives are structured way differently than what would be in a small business.
The BigCo can actually pay you big if you can improve the career and boost the ego of those who will make the decisions. To do this, you need to give people something that they can present to their own managers in a very positive light so they can present it to their own managers. That's how you can build a momentum of people who can push for purchase of your product.
From what I'se seen, making the life easier of some technical employees is not good. Think putting something trendy in it that the managers would love to boost about it. Could be a trendy tech(big data, AI etc.), could be social proof(Apple and Nvidia are also buying from this company). Just try to structure your marketing(which roughly means finding your customers, not to be confused with advertising) in a way that matches the incentives of people who don't spend their own money but try to work their way up in structured relationship. Of course, your product should also accomplish something so you can have a long term business. So it's not a fraud, it just needs right kind of engagement.
I agree - and doing this is not the same as doing everything they ask, on the bet it will come off. In my opinion having that view for the managers to boast about means you are aligned with where they will invest - it is where they see value.
> Some of our customers still think we’re one of those big enterprise software companies where you call them up, negotiate with a salesperson for three months, and, on the last day of the fiscal quarter, force the salesperson to promise a long list of new features in exchange for a half-million dollar contract.
> That’s nice, but we don’t have salespeople and we’re not one of those vendors. Our customers are happy that our price point is low but not all of them quite understand the implications. They ask us to fly out to their headquarters to give a demo of the software to their development team. They send us long spreadsheets with lists of features and ask us to check off the features we support. They even send us RFPs (shudder). RFP stands for “Request for Proposal.” It’s a request by a large company for a custom proposal from a small company. The small company works on the 200 page laser-printed proposal like mad for three weeks and Fedexes it in great expense and at the last minute, where it gets put in the trash because the large company has their favorite vendor who takes them on a helicopter to Atlantic City on junkets involving blackjack and strippers, and who is going to get the contract no matter what, but someone in purchasing for some unexplained reason, maybe he’s bucking for a promotion is insisting that the proposal be opened up to “competitive bidding” and the small company has been chosen as a victim to write up a proposal that has no chance of being accepted just to make the process look a little bit less corrupt, and if you’re a small company, I would recommend that you don’t fall for it and don’t spend any time responding to RFPs unless it’s already understood that you’re going to get the contract.
Joel, if you're accidentally around, your archive is borked :-( The front page reading lists are broken and navigation is super hard to try to find some of the awesome old articles.
I agree. For many BigCos this is not an active intent of malice. Often the 2 people do hope their vision will change the company. However, the company is often filled with these groups of "2 people" all trying to move the needle also, leading to a single vision not being effective.
My advice is have an honest and frank conversation with BigCo. Let them know it is in no one's interests for your start up to go out of business, and when they are doing at things at scale you (startup owner) can discuss passing on those economies of scale in terms of dedicated work, discounts, etc.
Fivetran has been diligent the three years I was tucked away in FAANG analytics. When I emerged earlier this year back into the world of normal companies, I realized there was a sort of holy trinity of companies that has completely dominated the analytics SaaS market: Fivetran, Snowflake, and dbt. I had only heard of Snowflake before.
tangentially related: if I'm already using Fivetran, DBT and Snowflake, where does Databricks fit in? From reading their marketing material, it would seem that Databricks is a full featured solution that replaces all of the above and offers everything from ingestion to ETL and even analytics. However I keep hearing of companies using Databricks alongside those other products and I don't understand how does it fit into that type of data pipeline.
Databricks is more of a platform than "just" a tool. So yes we offer first-party "end to end" solutions, and we're happy to show their value to you and convince you to pay for them, but we play nice with others, and we have a ton of great partners who can connect into our platform and help you ingest, transform, analyze, and consume your data. So you can use our native ingestion tools, or partners like Fivetran; our native ETL/ELT tools, or dbt; our native SQL/BI/warehousing capabilities, or Snowflake.
The metaphor Andy Kofoid our VP of Ops uses is XBox or Steam. They're a platform others can build things on, but they also sell their own games and offer value-added services on top of the things others are selling.
And like those platforms, we have the same ultimate twin goals of attracting customers and partners.
Its a great stack and works well together. Ingest with Fivetran, transform with DBT and store in Snowflake. Simple, serverless, consumption based pricing etc. definetly better than the old ETL world!
I remember when Fivetran came out years ago and it was a big improvement over the competition at the time. It made a particular kind of work easier, so you would hopefully know about it if that’s what you did for a living. I’m in a somewhat different speciality now so no idea how things have developed since then.
I almost joined fivetran as a staff engineer, I had a terrible experience in the hiring pipeline - the manager i spoke with was convinced he was sounding clever by asking someone with 10 years of experience the difference between a java hash table and hashmap. Also the scope of work offered was at the level of a early senior engineer, i felt the vibe this company stagnated 3 years ago.
Fivetran created a great mid-market product that was best in class and had wide adoption. What they didn't have was a very customizable product designed for large enterprises where their one size fits all strategy doesn't work. That is what the acquisition was about, getting that product so they could serve both markets.
Amazing that they could get $500+mm at a valuation of $5+ billion to buy a company with revenue of $30mm. Times have changed since then! I wouldn’t look at this as a guide for today’s market.
Is this a rags-to-riches overcoming-the-impossible story where the protagonist is already unfathomably rich and well connected at the outset and could buy their way to success? Not sure how much I can relate to that, tbh.
As soon as I read "whose family also summered in the same patch of northern pines" I knew where this story was going.
Basically read as: we started a failing company then used family connections to raise a fuckton of money to buy a successful company, and now we're successful.
I'm sure these founders grew up at least middle class, but you definitely don't need FU money to summer in the northern pines of Wisconsin. It's not exactly the Hamptons :)
These guys apparently had spent $160 million and according to the article, had “no product here.” It’s perplexing to me that they were then able to so easily raise hundreds of millions of dollars to buy competitors that did have a successful product.
And this was barely a year ago. Why do investors believe two failed founders will be able to grow and scale this business when they couldn’t grow and scale their own business?
I’ll personally make sure to avoid this company as I don’t at all respect their method for creating a viable “business.”