Code Brain | Where Do Great B2B Startup Ideas Come From (Part 2)

Through several founder stories, we analyzed how to find great B2B startup ideas from "past pain points" and "thinking and exploration". Today we'll explore the third path: current demand.

Current Demand: Identify the parts of what you're currently doing that show signs of demand, then go all in. The final path is to continue with your current product or business, but pay attention to features or characteristics that existing users strongly need. This might require minor adjustments, or it could lead to a product overhaul. In Amplitude's case, other founders wanted the internal analytics tool that the team had built for their own product — and that became Amplitude's startup idea.

"Before Amplitude, we founded a company called Sonalight. It was a voice recognition app that let you send and receive text messages by talking to your phone. We realized that to improve the product, we needed to clearly understand what users were doing in it. We needed to know where users were getting stuck; what made them stay; what they liked and didn't like.

We tried many products on the market. I remember Flurry, Google Analytics, Adobe, Kissmetrics, and others, but none of them met our needs, so we had to build our own. We were a bunch of arrogant engineers, haha. We ended up building an analytics tool, and when we shared these insights with other companies we knew (we were in Y Combinator at the time), they said, 'Wow, this is amazing. I really need to understand my business.'"

— Spenser Skates, co-founder and CEO

Amplitude is a user analytics platform that tracks and analyzes user behavior in apps and on websites. It features a simple, easy-to-use interface, flexibly supports various data sources and analytics tools, and scales to meet enterprise needs.

Sprig's story is similar — Ryan Glasgow discovered that a side project he was working on had more demand potential than his official business. "The startup idea I was exploring wasn't producing the results I expected, so I started doing user research from my strength in product management. I would ask tough questions like 'Would you pay for this?' or 'Why do you use this product?' I wanted the research process to be as rigorous as possible, because the decisions I made based on it carried huge risk (I might spend ten years on this). I discovered a concept called 'Outcome-Driven Innovation' that could make product innovation more rigorous. I got really excited about it — what if I could build a tool that let people validate ideas without building anything?

So I developed an SDK that could be integrated into apps. The interface was very clean. It would ask users rotating questions about functionality, usability, and quality. I used Airtable for the backend — very simple.

I shared this little tool with other founders who wanted to move their businesses forward, and gave them scorecards for quality, usability, and functionality so they could analyze using my tool. The feedback was great — people said they learned a lot from it. They even said they'd be willing to be the first paying customers. That's when I realized this was the market demand I was looking for, and I quickly threw myself into it."

— Ryan Glasgow, co-founder of Sprig

Sprig provides a suite of tools and resources to help users systematically think through, evaluate, and test ideas and concepts, thereby improving their success rate.

Before founding Loom, the founders had built two completely different products. The idea for Loom came from an unremarkable little feature in one of those earlier products. "Loom was the result of two failed video startups before it. Our first project was helping companies hire domain experts (like product managers, designers, engineers) for consultations on checkout flows, interactions, sales funnels, onboarding, and so on. We had some revenue, but not enough traction to justify continuing to invest time (we'd already been at it for four months). The second project was a SaaS tool that could be integrated into company websites to collect feedback through user testing. Similar to the first product, after three months we couldn't find a reason to keep going, though it did give us some genuine insights from design partners.

The highlight of that second project's three-month lifespan was that one client recorded a video summarizing all the user feedback their team had collected. We realized there might be something there.

A month later, we launched the product on Product Hunt and got thousands of downloads on the first day. That confirmed we should go all in on this area and stop looking in other directions."

— Shahed Khan, co-founder

Loom is a video recording and sharing software product that helps people communicate and collaborate more effectively. Loom supports screen recording and webcam capture, and lets users add text, images, and audio. Loom also provides a range of tools to help users edit and share their videos.

Segment tried many ideas over a year and a half before one finally worked. "We were still college students when we applied to YC. Our proposal was for a classroom teaching tool for universities, which we developed over one summer semester. Because we weren't good at finding real demand, the project failed. We thought one reason was that we didn't understand our product's users.

We spent 15 months building various analytics products to better understand users, but still nothing took off. We kept trying new directions to attract users. As a growth hacking tactic, our co-founder Ilya built a small library called Analytics.js that users could use as a plug-and-play tool for Mixpanel or Kissmetrics. Through Analytics.js, you could connect one library and send data to each of these platforms separately.

With only six months left in the batch, we still hadn't launched a product. My co-founder Ian saw big potential in the demand behind Analytics.js. We could become an efficient API layer among all these boring, similar-but-not-identical data tools.

We disagreed on this. Peter thought it was the worst idea he'd ever heard — you couldn't build a company on 100 lines of JS. Eventually we agreed to spend one week building the product and launch it on Hacker News. On launch day, it hit number one on Hacker News, and nobody had any more doubts."

— Calvin French-Owen, co-founder of Segment

Segment is a customer data platform that helps businesses collect, manage, and analyze data from different sources. It can connect to various data sources, clean and standardize data, and send data to other analytics platforms.

Slack's origin story — its idea coming from a popular feature in a failed video game — is already well known. "We all agreed that the game we were developing, Glitch, would never become a commercial product worthy of its $17.2 million in funding. If this was something six people had built for a million dollars, that would be fine. But by late 2012, we had 45 people and had spent several million dollars. The game's user base wasn't going to grow further, so we decided to shut it down, and we didn't know what to do next.

We still had several million dollars in the bank. We could return the money to shareholders and call it quits, or we could do something else. Because we still had funding, we could wind down the company gracefully. We spent a long time helping employees find new jobs to make sure they wouldn't be unemployed.

We also gave players the option of refunds — they could leave the money with us or donate it to charity. We were actually quite busy during the shutdown period, and we were also thinking about what to do next. There were all kinds of ideas.

It took us some time to settle on a direction, which eventually became today's Slack.

This came from how we communicated during game development. At the time we used very old-school IRC. IRC could only do so much, and we gradually added little features we wanted.

For example, on IRC, if both people weren't online at the same time, I couldn't message you — I'd have to wait until you came online. So one of our first improvements was making messages archivable, so when someone came online they could see what others had sent while they were offline. We wanted to be able to search these archived messages, so we added search. There was no good iPhone client, so we built an HTML5 archive viewer.

Something interesting happened. Until we shut down the game, the company had 45 people, and we'd been operating for three and a half years. During that entire period, the company-wide email list sent only 50 emails — roughly one every three weeks. This wasn't intentional, nor was it a conscious decision to avoid email. When everyone checked IRC first, people increasingly valued IRC, and the more they valued it, the more information we put on IRC, and the more information on IRC, the more people only looked at IRC. Eventually, everything from database alerts to daily sales data was announced through IRC. Even whenever someone uploaded something to the file server, that notification went to IRC. We didn't succeed at making games, but our process of making this unsuccessful game was incredibly efficient."

— Stewart Butterfield, founder, via Business Insider

Summary and Reflection

• Among all the features you've built, which shows the clearest demand?
• Is there a problem at your current startup that you wish someone would solve? If so, try building the solution yourself.
• What side projects do you want to do but haven't started yet?

Three Paths

To summarize, if you want to find a good B2B startup idea, it should have these qualities: 1. Important; 2. Poorly served; 3. You're interested in solving it. You can choose one of the three paths we've outlined and think through the corresponding questions:

Past Pain Points

• Did any of your past work or products you've built have unique value for a company or colleagues?
• What tools did you especially want in your previous job (and how much would you have paid for them)?
• What product did you have to rebuild from scratch at every company you've worked at?

Thinking and Exploration

• What trends are emerging but poorly served (e.g., security, collaboration)?
• What commercially viable technologies are emerging but undervalued or underdeveloped (e.g., data scalability)?
• How many potential customers have you discussed your startup idea with?
• How important do your potential customers think the problem you're solving is, on a scale of 1 to 10?
• How poorly served is the current market, on a scale of 1 to 10? How would you rate your solution? It needs to be at least 4 points higher to have a chance.

Current Demand

• Among all the features you've built, which shows the clearest demand?
• Is there a problem at your current startup that you wish someone would solve? If so, try building the solution yourself.
• What side projects do you want to do but haven't started yet?

When your current project or work touches on "pain and demand" in your industry, you'll naturally sense its potential. If you've always wanted to make something happen in B2B, use this framework as your foundation and look for your opportunity in your past experience.