Everything is still BI

What to expect when you’re not expecting to build a BI tool.

It’s hot, but sure, touch it if you want too.

Look, I get it. You've worked around data for a while, and seen a better way to do things. You worked on a finance team that finally figured out how to do revenue reporting for software-as-a-service businesses. Or you’ve consulted for 50 different ecommerce brands and know the standard set of metrics that everyone needs to measure marketing performance, or you’re a data visualization expert who’s frustrated by how dashboards strip the narrative away from the numbers. Or you’re an analyst on a data team somewhere, and you’ve built some internal data tool that finally gets right what all the other tools in the market get wrong. 

So you’ll start a company. You’ll make fundraising decks that tell people that you’ve been in the trenches, and you know that the hard part of working with data isn’t making a chart, but analytical reasoning—and today’s BI tools don’t solve that problem. 

But your tool will. You aren’t made for generic business users like the geriatric business intelligence tools of generations past, you might say. We’re something new; something more specialized. An A/B testing tool that finally works; Salesforce reporting that isn’t terrible; a collaborative data platform. You’ll say you were built for experts; built for sales ops teams; built for analysts; built to make data scientists more productive. You’ll say that BI is a broken promise, a gimmick, shelfware; we’re bottoms-up, user-driven, consumer-grade. You’ll say something about turpentine.¹ 

“Unlike traditional BI products,” you’ll say in investor pitches, “we help analysts share high-impact insights with greater speed and flexibility.” People will like the pitch; they’ll buy into your story. “All the other analytics companies that I’ve seen are building dashboards of one sort or another,” an angel investor might say. “There’s definitely a market for that, but they’re doing something different.”

BI is constrained; we are flexible. BI is stale; we are fresh. BI is so irrelevant that you won’t just put BI vendors in a different hemisphere of your mandatory two-by-two competition grid; you’ll leave most of them completely off the map. Your real competition, you’ll say, is people not realizing that there’s a better way, and SAS.²

The market is ready for the revolution, you’ll say. You’ll have a “Why now?” slide in your pitch decks, and it will be filled with facts that only a fool could disagree with. Ten years ago, you might’ve said that data science is going to be the sexiest job of the 21st century, that data scientists’ workflows are broken, and that they need dedicated, purpose-built tools of their own. You might’ve said that Redshift would make analytics more accessible. You might’ve said that building a multi-tenant SaaS platform gave us an economic edge over the traditional on-premise vendors. You might’ve said that you’re riding a “strong trend toward ‘democratization of data.’” Today, you might say that analytics engineering is going to be sexiest job of the 21st century; that centralized data platforms like Snowflake will make analytics more accessible; that using AI for development gives you an economic edge over traditional software vendors; that you’re riding a strong trend towards democratization of generative AI. There’s a new gold rush, and you’re selling shovels. 

Some people might ask you which Gartner Magic Quadrant™ you’re in. Ridiculous, you’ll say. The old categories were for Qlik, and Spotfire, and Oracle Business Intelligence Suite Enterprise Edition Plus. We’re something that can’t fit into the categories of yesteryear. Crazy one today, category creator tomorrow. In your sales decks, you’ll cite Gartner’s own research to prove your point. 

Your first customers will tell you that you’re right.³ Your product “is transforming entirely the way we approach data as a company,” a head of product will say. It “freed up resources to tackle new types of questions,” a vice president of analytics will tell you. You’ll hit milestones. Hundreds of customers; past $1 million in revenue; past $10 million; past the point of supposed inevitability. You’ll start hiring more sales reps; someone from another data company will join as your senior director of alliances. You’ll talk about selling to the enterprise.

But—you’ll feel a nagging anxiety. You raised venture money, and venture capitalists want exponential growth. Your growth is more…exponentialish. It was 150 percent two years ago, and 100 percent last year. You’re forecasting 80 percent this year, and that’s assuming you make up for your slow Q1 in the back half of the year. 

You have a plan, though. We can make money from those generic business users. Because your brand is a power tool for power users, you’ve let the masses use your product for free. But it’s been so successful with the specialists they’ve started sharing their work with the rest of the company. They now make up a majority of your user base. 

And they’re asking for features. You’ll lose a promising deal at Wayfair because their marketing team preferred Looker. Your most requested feature will become more visualization types. Then reports, delivered via email. Then better dashboards, and easier ways for people to access data without writing code.

You’ll tell yourself that you can do these things without losing your identity. You’ll be BI, for people who love spreadsheets. You’ll be BI, for Shopify stores. You’ll be BI, reinvented for the age of Slack; for the age of dbt; for the age of the data lakes; for the age of AI. You’ll be notebook-powered BI, or SQL-powered business intelligence. You’ll be an SUV, with the soul of a sports car.

Other people, however, won’t see it that way. For prospects, your original brand will be a chronic condition. They’ll still see you as a spreadsheet, a notebook, a SQL workbench, or a reporting tool for Shopify stores.

But for customers, the new brand will be a betrayal. They had dozens of BI tools to choose from, they’ll say; we bought you because you were different. I’m surprised you built that feature request; now we have a renewal to address. 

Internally, you’ll debate what to do. Are we a BI tool? Should we embrace it? You’ll ask someone to do a market analysis. Or should we go back to our roots? You’ll never quite decide. You’ll experiment; flirt with the idea; “maintain optionality.” You’ll tell yourself you’re still committed to your original users, but your roadmap will be full of the features that look a lot like PowerBI. Your messaging will change depending on who you’re talking to. Your brand will become some useless catchphrase. 

You’ll doubt your past decisions—what if we had never built any BI features at all? What if we just kept building the thing people loved us for in the beginning? What if we raised less, and focused on product craft and profitability? 

Ah well, you’ll think. Maybe next time, you’ll do it differently. But this time, the venture checks have been cashed. The roadmap has been promised. Even though you told yourself you weren’t going to build a BI tool, that’s where all the money was. So you sailed towards the storm, and now, she’s not going to let you out. 

---

Anyway, last week, Equals, a cloud spreadsheet tool that connects to databases that we’ve talked about before, launched Equals 24. It includes more visualization types, better dashboards, and easier ways for people to access data without writing code. From their release video:

We've made a ton of progress towards solving some of the biggest areas of feedback we've received over the years. ... 

[There is] an all new query builder [that's] particularly powerful for people that don't know SQL. Of course, if you do, Equals supports writing SQL; you can write as complex a query as you'd like. ... We've completely overhauled charts based on your feedback, of which there was a lot. New chart styles, colors, and customization options make it easy to create beautiful charts... [And our dashboards] look better than ever now with those new chart styles, and we have a bunch of design improvements coming very soon, including better-looking tables, layout options and spacers.

That's Equals 24 in a nutshell. It's still a spreadsheet at its core, but not just a spreadsheet anymore.

What’s striking about this release is that we said the exact same thing at Mode. Our core was SQL tools for technical data scientists, and Equals’ core is spreadsheets for finance teams—and both were well inside the event horizon of the BI black hole. Better dashboards, more visualization types, and easier ways to access data were, quite literally, our top feature requests too. 

It’s the gravity that none of us can escape. ThoughtSpot’s initial ambitions were to “eliminate the need for traditional BI tools through ThoughtSpot Relational Search Engine;” today, it’s thoroughbred BI. Hex, which started as a collaborative notebook for data analysts, built dashboards and drag-and-drop visualizations. Transform, which started as a semantic layer, built dashboards and drag-and-drop visualizations. Narrator, which started as a data modeling product, built dashboards and drag-and-drop visualizations. And the SQL chatbots are building dashboards and drag-and-drop visualizations.

In other words, no matter when a data product was created or what its first features were, it will inevitably collapse into the same singularity: Database connectors; a code-based query editor; a code-free query builder; lots of charts; dashboards; self-serve.

---

Still, I get it. You’re not convinced. You’ve heard that “everything is BI” before; it’s my whole bit.

I’ve been there before too. During one of Mode’s early fundraising blitzes, we were approaching the finish line with a prospective investor. Our sponsoring partner was in favor of doing the deal, and we’d been asked to make a closing pitch at the home of one of the firm’s senior leaders. The partner, a founder of a $50 billion public company with a house in Sea Cliff to match,⁴ was our final boss.

We got destroyed. Within the first half hour, we all knew the deal was dead. The next 90 minutes were polite theater. As we were leaving, we all shook each other’s hands and talked about “next steps,” though everyone knew that there wouldn’t be any. The partner left us with a final bit of advice: We'd never make it without building an in-memory database. We were a BI tool, he said, and BI tools need fast, in-memory compute engines.

On the car ride home, we rolled our eyes at his dumb warning. Ok boomer, you irrelevant fossil. We weren’t a BI tool; we were something new; if he couldn’t see that, we didn’t want his money anyway.⁵ 

And then, six years later, we rebuilt Mode around an in-memory database and rebranded it as Modern BI.

There are a lot of lessons to take away from this story—most of them about hubris, some about technology, and one about never turning down wine from a billionaire’s cellar. The biggest lesson, though, is that it's actually very easy to think different, but it’s very hard to stay different. 

When people talk about Silicon Valley’s icons—people like Steve Jobs, Jeff Bezos, Elon Musk, and Sam Altman—they tend to describe them as visionaries. They built huge companies and personal fortunes because they saw something the rest of us didn’t, and were smart enough to figure out how to build it. I’m not an NBA star because I don’t have Russell Westbrook’s nuclear athleticism or Victor Wembeyamba’s alien proportions. I’m not buying Hawaii because I don’t have Mark Zuckerberg’s big brain.

But in truth, that’s probably a cope. These people are successful for a far more accessible reason: They’re stubborn. Most of their ideas weren’t all that new or profound. Instead, they were uniquely insistent on what they wanted to build, and uniquely unrelenting in building that thing. Steve Jobs didn't care what customers thought; Jeff Bezos didn't care what his employees thought; Elon Musk didn’t care what the U.S. Security and Exchange Commission thought; Sam Altman didn’t care what Scarlett Johansson thought.⁶ It is not having a vision that gets you labeled as a visionary; it’s refusing to let it go. And that’s something any of us could do, but very few of us actually do.

Similarly, you could build something that’s not BI, but you won’t do it by being clever. You’ll do it by being stubborn. Customers will demand dashboard and drag-and-drop visualizations. Your market analysis will say that’s where the opportunity is. Your financial models will look better when you can sell to generic business users. Bit by bit, building a BI tool will become the “smart” choice. If you want to be different, it takes will, not wits.

1. When smug tech people get together, they talk about talking about turpentine. 

2. Of course we had one of these slides too.

3. No selection bias there. 

4. The entire backside of the house looked out over the mouth of the San Francisco Bay, with a full view stretching from the Presidio to the Golden Gate Bridge. When we got there, the partner apologized that the windows hadn’t been cleaned in a while, and assured us that he’d be moving into a more respectable house soon.

5. lol no we very did. 

6. To be clear, I’m not saying this is good, or ethical, or legal. But it does seem to be fairly effective.