Tips on growth marketing

February 07, 2019

I recently listened to Julian Shapiro’s interview on Indie Hackers, Things Every Founder Should Know About Growth. Julian runs Bell Curve, a growth agency for tech startups. In this podcast, he gives a wide variety of tips on growth marketing. I’ve taken notes here.

What kind of company should you start?

Controlling for founder skill, e-commerce has the highest rate of success. Afterwards, mobile apps (particularly those that deal with transacting finance, like friend-to-friend and Bitcoin), and then Chrome extensions. At the very bottom are SaaS businesses. Basically, it’s actually somewhat hard to make software that people want.

Why is e-commerce more successful than SaaS? It’s easier to convey the benefits of something like a toothbrush than it is to convey the benefits of your customer support software. People have short attention spans, so if you need to take half of the ad to explain why people should want your product, you’re stacking the deck against yourself.

In addition, people are already going to want to buy a toothbrush no matter what. You can just help them make the decision to buy yours. It’s also visually easy to understand. And finally, there’s a price in mind already: people know that a mattress is hundreds or thousands of dollars, but SaaS apps don’t have a set-in-stone value, so people are more price-sensitive.

Some people are intimidated by the idea of creating a physical good and shipping it across the country. In practice, it’s easy because there are full-service providers who will handle distribution for you. (He points out that LA-based e-commerce startups have a very high rate of success due to the networking effects.)

Developers think that more software is the answer. It’s not always the case.

How do you choose a business idea?

Julian’s advice on choosing a business idea is rock-solid:

  1. Determine your target income.
  2. Which of your ideas are likely to achieve that kind of income?
  3. Out of those ideas, which are you most passionate about, and which of them solves a problem that you have?

The second question in #3 basically qualifies that you understand the problem well; it also enables you to be your own customer, which gives you a better view of the landscape.

Find out if your idea has growth potential: that is, if it will be able to sustain the financial lifestyle you want without too much friction. To do this, look at what type of business you’re running (like I wrote earlier, e-commerce has the highest success rate), review the market size, and learn how badly the market wants what you’re offering.

The best way to know if the market wants what you’re offering is to see if you have competitors that are doing well. If they are, this is a very good sign! You can measure your competitors’ performance by proxy:

How do you validate an idea?

Ads are the quickest, most controllable, most reliable form of getting traffic to your website. Word-of-mouth and organic traffic can take months or years, if ever.

So you can use ads to validate ideas. Put an ad on the top of Google to see how many people will click through; put an ad on different social media platforms to see how much engagement you get. This helps you determine if there is demand for your product.

If you want to try your hand at unpaid advertising, do keyword analysis on terms in your industry, and write content to catch those people. Develop a good call to action and see if they sign up.

What are the best channels for ads?

If you are selling to businesses, you may have to rely less on ads and moreso on traditional sales. Nonetheless, here are his recommendations for a few types of businesses:

For B2C e-commerce businesses:

  • Instagram
  • content marketing
  • a little bit of PR to kickstart organic efforts
  • potentially Pinterest and Google AdWords

For B2C SaaS businesses:

  • Facebook ads
  • content marketing
  • potentially Google AdWords and affililate programs

For high-end B2B SaaS business ($1,000/month/customer):

  • sales
  • cold emailing
  • traditional networking
  • cold LinkedIn messages
  • lead generation through ads on Facebook and Google

The general thought process on each of these are:

  • Do my customers spend time there?
  • Are they in the right frame of mind to make a buying decision?
  • Is it too expensive to run ads there, given how much I make from my average user?

Later, Julian says that the channel doesn’t matter as much as the person you’re targeting does. Julian’s best performing ad for a company that sells $50,000 machine learning computers is an Instagram ad of a computer with doggy ears on it. As long as you’re targeting the right person, it doesn’t matter where you’re targeting them.

How do you position a product for growth?

If your product handles a lot of use cases, use data to find the use case that is 1 most common, and 2 relieves the most pain. This has the best chance for explosive growth via targeted ads.

Some tips on this:

  • is what you’re doing successfully solving a problem?
  • can it be conveyed succinctly and effectively?1
  • is the signup process relatively frictionless?
  • can you get results for customers before they pay you?

Any tips for B2C e-commerce?

Identify e-commerce trends where branding is the differentiator. For instance, by looking at third-party Instagram data, you can see what is being talked about the most. If you see a trend that is growing out of nowhere, you may want to jump the gun early and be there from day one. Assess your idea by going to Google Trends, enter the name of the trend, and see if it’s spiking up in unison. Then, focus on creating a company in that niche and differentiate yourself as the “go to” product with your branding.

Recently, Julian has seen companies that have succeeded doing only the following: finding a product that is old and overplayed: mattresses, toothbrushes, etc. Making a beautiful, modern brand around it: designing it better, making it customizable to the particular customer, making the brand more chique, and painting a story around it. These companies trap people in a bubble. While companies succeed doing only this, Julian views it as potentially bad for the marketplace. If you’re a branding expert rather than a developer, this is a good option available to you.

How do you target users within a niche?

You can target Twitter users based on who they follow. So if you’re trying to sell a nutritional supplement, you can target people who follow industry leaders in nutritional supplements. You can also target people who’ve entered nutritional-supplement-based keywords into their tweets. The downside is that Twitter is one of the worst ad channels in terms of cost-per-click – worse than Facebook, Instagram, and even Snapchat. In addition, SEO is important for niche users. Find out what your niche users are searching for, and then dominate that niche. Finally, Facebook can also do lookalike targeting: they look at the people who are visiting your site and promote your ads to people who have similar behaviors.

How can you build a viral product?

How can you find emerging advertising channels?

Any time a social network is gaining traction, it’s almost a given that they’ll introduce an ad platform to monetize themselves. Set up Google Alerts about “[platform name] ads”, and get on as early as they can.

As of February 2019, Quora is emerging, and Snapchat and Pinterest are beginning to perform better. Periodically review older platforms, as well, because they have a lot of employees trying to improve the experience of their platforms since it’s their primary revenue driver. Spotify and Pandora are also releasing ad platforms, and Julian says to give them a whirl.


  1. The best ad Julian’s company created for their client Service read, “PS. Airlines will pay you up to $135 when your flight is delayed or cancelled.” There is no other text on the ad, but it performs so well that they’ve spent seven figures running it. ↩︎

  2. According to Julian’s growth marketing guide, videos are good for ads, but not for the hero on your homepage. Users won’t watch your video because they think it will take too long to get to the point, and then they’ll skim your content as fast as possible to learn about your product (which leads to high bounce rates). Use a screenshot instead. ↩︎