How to save yourself from a bad startup idea that looks good


The startup bug has bitten you. You want to start a business, grow it for a few years, sell out and rest easy for the rest of your life. A great dream to have. But that’s the easy part. The hard part is building the business. And this long, arduous journey starts with a single step – having a great idea.

How do you come up with a startup idea? To start, you read this article by Paul Graham of Y-Combinator. It’s thought-provoking, even by Paul’s lofty standards. Paul says a lot about the characteristics of great ideas. But he also talks about a similar-looking but antithetical concept – the “sitcom” startup idea.

What is a sitcom startup idea? It’s one which sounds plausible, but is actually bad.

This is not just a bad idea. We have tons of those, and they are easy to identify. Even if our ownership of the idea blinds us to its infantile stupidity, our friends will warn us. They’ll tell us it’s the dumbest thing they’ve ever heard. And we can swallow our pride and move on to the next idea.

No, the sitcom startup idea is not bad in the same way. It’s an idea which sounds plausible. So plausible that when you go ask customers whether they’d use it, they don’t say no.

This is what makes it dangerous. You can read Lean Startup, dutifully ‘validate’ your idea with customers, and then build it. Only to find out that there actually is no market.

Social network for pets

Paul illustrates this with an example of a ‘social network for pets’. If you have pets, this sounds like a good idea. Sure, you can imagine posting photos of your pet parakeet on, where others are waiting with bated breath to “like” them. Or, what’s far more insidious, you can imagine others around you loving this service.

I actually tried this during my lecture in IIM Trichy, and people loved the idea. But it’s bad on two levels:

  1. It’s erroneous to assume that if people say they like a product, they’ll use it. I might like 30 different websites, but that doesn’t mean I’ll check all of them every day. Given my limited attention span, the only social network I’ll use daily is Facebook.
  2. If you talk to 100 people and they all say they “know someone who would use this”, then you’ve found yourself a community of 100 almost-users. Or to be precise, exactly zero users.

So how do you differentiate between sitcom startup ideas, and truly promising ones? How do you know if you’re on to something huge, or just a mirage?


The short (and hard) answer is – you try anyway. You build an MVP and check if there’s traction in the market. If there is, congratulations, it worked. If there isn’t, then you know you just had a “sitcom” idea.

But there is an easier way. I’ve come up with a few patterns to identify what is probably a bad idea, even though it sounds plausible.

Before we jump in, a caveat. I don’t know if any plausible sounding idea is actually bad. What I do know though, is that the universe of plausible ideas is much, much larger than the set of good ideas. So, an idea that is only plausible is probably bad.


Just like I know that a monkey banging away at a keyboard will not produce Romeo and Juliet (it might, but the probability is infinitesimal), if all I know about a startup idea is that it’s plausible, it’s probably bad. Sure, you get a Twitter every once in a while. A product that seems random can suddenly catch fire. But such instances are so few and far between that you can ignore them.

With that done, let’s dive in to the patterns:

1. Broad and shallow, vs. narrow and deep

One of Paul’s theses in his article is that you should solve a deep need for at least a few people. If the need you are solving is shallow, then it’s not a great startup idea. Even if it affects a broad set of customers.

It’s got to be a major problem – a mild or one-time issue won’t cut it. You’ve got to create a product that at least a few people NEED, not one that a large number of people WANT.

A sitcom idea of the ‘broad and shallow’ variety can follow several patterns.

A “vitamin”, not a “painkiller”

The social network for pets falls into this category. It’s a nice-to-have, like a vitamin capsule. No one needs it, like the root-canal patient who’ll pass out without a painkiller. If people just ‘want’ what you’re building but don’t ‘need’ it, tread with caution. You may be onto a bad idea that sounds good.

Not solving a top-tier problem

But only solving a problem is not enough. It has to be important. Simply put – if the problem you’re solving is not one of your customer’s top 3 problems, it’s not important. Give up now, before it’s too late.

I once thought of building a software tool to help VCs manage deal flow. It would have a visual funnel, to tell the VC how many deals they have seen in the last 3 months, and at what stage of discussion each deal is. And they could dice it by any filter (e.g., SaaS vs. consumer, location, stage of business, etc.) to see their deal pipelines.

A great idea, I thought. The only issue – it’s not an important enough problem. Getting strong deal flow is far, far more important than tracking it. Many VCs are happy enough using Excel to track their pipelines. They’re not even trying generic funnel management systems like Salesforce. Why will they bother using one tailored for VCs?

“Solving a problem people don’t know they have”

This is a first cousin of the two patterns above. While not a “vitamin” solution per se, it’s solving a problem people don’t know they have. Which begs the question – how do you know they have this problem?

I tried doing this a couple of years ago, with a plug-and-play loyalty program for small business websites. Users would get points for coming back to the website every day, reading articles, sharing to social networks, etc.

A great idea for large, stable businesses trying to increase customer retention, maybe. But a small business finding its feet? These guys don’t even think about gamification or loyalty. They have other problems. They need to build a user base first, before trying small tricks to engineer loyalty.

I tried selling this for 6 months. It did not work. It’s hard enough convincing people to buy your product. Why do you want to add the burden of convincing them that they need it?

UPDATE: Mike Fishbein makes a similar point in this article. If you want to avoid building something no one wants, then solve known problems.

“This product solves everyone’s problems” OR the “Microsoft Office” product

I love Microsoft Office. It’s so flexible, so all-encompassing. No matter what type of problem you’re working on, you can bet that Excel and PowerPoint will be super helpful. Or think of Google – no matter your query, you can find the answer.

These are all excellent products. But aiming to solve everyone’s problems in one go can sound the death knell for startups. Why? Taking the example of my gamification system again:

  1. It’s unlikely that there’s a dire need for your product among a huge mass of people already. If you’re solving a problem for everyone, it’s probably a broad and shallow problem, not a deep one. My system was a nice-to-have, not the answer to their top 3 problems.
  2. In most cases, flexible products necessitate a learning curve among customers. Newsflash – your customers are too busy to spare any time to learn how to use yet another product. Unless you’re solving a problem as critical as the ones Office and Google solve, good luck getting adoption. It’s more sensible to focus on one type of customer, and solve their problem better than anyone else.
  3. Solving everyone’s problems at the same time requires a complex back-end. Why build that without strong market validation first? You’ll either end up building a buggy product, or worse, build a great product that no one wants. In the case of our product, the tech challenges proved intractable. Trying to integrate our system with several website technologies meant that it didn’t work well with any.

“Cool product I’ve built”

You get this a lot from engineers (I’m one too). We focus on the product, because we feel that the product alone is good enough. “My cool new app allows you to share your photos with all your Whatsapp groups in one go”. Great, but what if your users don’t want that?

“Build it and they will come” doesn’t work, in this world where a million apps are fighting for people’s eyeballs. See the chart in this article to see how high the bar is. You need to be sure that you’re solving a problem, and a top-tier one. Else, you could just be a “solution searching for a problem”.

Demonstrate need first. Else, your intricate product could just be another elaborately constructed pipe dream.

2. Templatized business models

“Uber for X”

[as used in “Uber for bicycles: On-demand bicycles for your riding pleasure”, or “AirBnb for cars: Rent other people’s cars when they’re not using them”]
"Do you want a bicycle at this very moment?"

“Do you want a bicycle at this very moment?”

This template is as old as the Internet. Take what’s working in one sector, and plonk it into another. It was “Website for X” in the 90s, and “Social network for Y” in the 2000s. But it’s a dangerous stratagem. Why?

Sure, Uber has been uber-successful in the cab market. But that doesn’t mean on-demand could work for every other sector. Unless the idea has grown organically from a problem, you have to assume it’s bad. You have to assume that the founder has applied the Uber template to the first sector he could think of.

Another clue that you’re facing this situation is when founders have no real expertise in the area they’re building for. Then how do they know that the problem is real? They don’t. All they know is that the solution is real, for another sector.

“X for India”

This is an even more pervasive and notorious template. Unless the model has some kind of geographical constraint (e.g., on-demand cabs), there’s nothing stopping a successful US business from expanding to India.

Moreover, if the model involves network effects, then you’d expect something that’s grown in one place to capture share rapidly in other places too.

As Mahesh Murthy is wont to say, the Facebook of India is Facebook. The Tinder of India is Tinder, and not Woo.

There’s one more problem with this template – some models just don’t extend across geographies. On-demand bicycles may be a great idea in Scandinavia or Taiwan. But it just won’t work in hot, sultry, noisy and overcrowded Mumbai (gosh, why am I still living here?).

3. Incremental business models

This is another type of business idea that we see quite often. It often involves just a slight tweak to solutions existing in the market. Again, this can be of two types:

Cloning an existing player, but with slight improvement

Think “Uber with wi-fi”. Of course, Uber has started doing this now. But even if it didn’t, this would be a horrible idea for a startup. Wi-fi is not differentiation. It’s a cosmetic touch-up engineered solely to help you raise money from rookie investors.

It’s wrong on two levels:

  1. It assumes that the incumbent will sit idle while you bring out an improved product. If what you’re bringing to the table is only an incremental improvement (i.e., 1x, not 10x), you can bet that the incumbent will also include it in their next release, if they find out it’s a helpful add-on. Don’t assume stupidity.
  2. Often, your improvement has nothing to do with the core problem you’re solving. Wouldn’t it be silly to say, “Uber works, but people hate the fact that it doesn’t have wi-fi.”?

Cloning an existing player, but in an adjacent market

Back when Bookmyshow (a movie ticketing website in India) was only a couple of years old, a friend told me he wanted to build a “Bookmyshow for Plays”. This is a bad idea too. Bookmyshow had already solved the harder problem of getting customers. So, it was much easier for Bookmyshow to include plays on its platform, than it was for a new player to start afresh. And true enough, plays appeared on Bookmyshow a few months later.

A giveaway for this kind of sitcom idea is a statement of the form “Today’s solution is satisfactory. But mine’s much better”. For your idea to be definitively good, today’s solution cannot be satisfactory! At least for a segment of the audience. Otherwise, your idea would be like my deal flow management solution for VCs. A nice-to-have, but not nice enough to change an existing process.

What is nice enough though, to change one’s existing behavior? A 10x improvement – whether in ease, time taken, or effectiveness.

4. “No Competition”

You often hear founders say that they’re the first team to do X, and that there are no competitors. Or they may say that everyone is a competitor (which is another way of saying “no competitors”). If you hear this, run in the opposite direction as fast as you can.

Why? Why is lack of competition alarming? For two reasons:

  1. If there really is no competition, maybe the market itself is unattractive. Today, it is difficult to come across a problem that no one has seen at all. Why do you want to solve unambitious problems, when it’s just as difficult as tackling ambitious ones?
  2. The founders may not have done thorough analysis, or may be suspended in the myth that their competitive moat is bigger than it actually is. Would you want to back such founders?

Wait, so am I saying competition is actually important? Yes – many players trying to solve a problem demonstrates strong need. But to succeed, you still need to differentiate. You need to have an ‘unfair advantage’ in startup parlance. Whether industry experience, critical partnerships, etc. – you must have a secret sauce in your recipe for success.

That’s it. Those are the patterns that should raise your suspicion antennae when listening to startup ideas. Am I missing any? Let me know in the comments, send an email to, or tweet at @jithamithra.

Of course, some ideas may actually be great, even if they fit these patterns. They may end up changing the world. Only one way to find out for sure – launch an MVP, and prove me wrong.

Related Posts: