Project: Craft a Referral Strategy

Phase 1: Setup

Make a copy of our virality spreadsheet.

Phase 2: Magic Moment

Phase 2a: Qualitative magic moment

Walk through your app or checkout flow. At what point does the user feel like they’ve gotten the greatest value out of your product?

This is when we'll ask for a referral. Write it down in your doc.

Sometimes it's obvious. Sometimes you have to ask users to know. We do this with user surveys.

Phase 2b: Data-driven magic moment‌

Another way to find your magic moment is by looking at your data.

It’s not always possible for SaaS companies to pinpoint their magic moment. For example, for Dollar Shave Club, the magic moment is right after someone shaves with one of their razors for the first time and sees themselves clean-shaven in the mirror. That can't be tracked digitally.

If your magic moment isn’t obvious, you can use a couple of proxies:

Proxy 1: Find out if people pay for your product again.

If someone doesn't cancel their subscription and doesn't buy from your competitor, chances are they felt your magic moment.

If someone finishes a free trial and decides to pay, that's a magic moment.

Use those kinds of pieces of evidence to zoom in on your magic moment.

Proxy 2: Cohort analysis.

A cohort analysis is a month-by-month breakdown of how many people stay with your app. It usually lives in your analytics tool (e.g., Mixpanel, Heap, Amplitude).

For example, say you see in your analytics tool that 40% of your users leave after a month, and another 30% leave after the second month. But 90% of users who make it to the third month stay at least a year.

The third month is the point at which most users stick around. You can usually see this graphically, without having to do any math. It's where the curve starts to flatten out.

Here’s a real example from Heap.

The retention point is clearly at one week. In fact, we’d even want to break it down by day at that point.

Can you find a retention point for your product? If you can, write it down. That's where you'll want to ask for a referral.

Phase 3: Word-of-mouth virality

Go through our word-of-mouth reading. Brainstorm three ideas that could apply to your product.

Add them to the word-of-mouth section of the template.

Phase 4: Pull virality

Go through our pull virality reading. See if it applies to your product. Remember, pull virality only works for certain types of products, and it's not something you want to force.

If pull virality is relevant to your product, add three ideas to the pull virality section of the template.

Phase 5: Push virality

Go through our push virality reading. Brainstorm a list of three ideas that could help increase exposure to your product. Add them to the push virality section of the template.

Phase 6: Incentivized virality

Phase 6a: Incentives

Go through our incentivized virality reading. Write down two or three incentives that you think would work for your product.

Remember to focus on value over cash.

Phase 6b: Ideas

Brainstorm three ideas for referral programs and incentivized virality. Add them to the incentivized virality section of the worksheet.‌

Think about what an incentive program would look like if it involved a real human being or a personalized touch—not just a templated email with a spammy-looking link.

Phase 7: Explore Viral Loops (optional)

Go to Viral Loops and sign up for a free trial. Follow their onboarding process, and poke around to see if any of their templates would apply to you.

Get comfortable with the UI. Walk through it and set up as much as you can. You don’t have to go live with it—just get used to what all the fields are and what everything means.‌

Phase 8: Prioritize‌

Go through each of your ideas. Rank them by three criteria: cost, time, and appeal. Use the second tab in this spreadsheet.

• Cost. How expensive it would be to implement. Try to get to a dollar amount. Take the salaries of the employees you'd need, figure out how much they cost per hour, and multiply that by the number of hours it would take. Add in the cost of any tools for the length of time you'd need to use them to test your referral program. (If this doesn't make sense, there's an example coming below. Hold tight and read on.)
• Time. How long it would take to get started.
• Appeal. How compelling the idea is on a scale of 1 to 10. Some ideas are just a lot sparklier than others.

See if anything floats to the top based on those criteria.

Example

Let's say you want to add a referral page to your app, with a link people can share with friends. You'll need to make the page using Viral Loops, plus your devs will have to hook in your API and set up your site to point at Viral Loops’ page at the right time. You'll want to test it for the length of your usual sales cycle and pull the plug if it's not profitable.

1) Cost

Dev cost per hour: $100

Dev hours: four

Total dev cost: $400

Viral Loops test length: two months (if that’s the length of your sales cycle)

Viral Loops price: $70/month

Viral Loops cost: $140

Total cost: $540‌

2) Time to push live (including QA and getting blocked): two weeks

Add a buffer of 1.5x for bugs, testing, and management.

3) Appeal

Pretty straightforward. Five out of 10.

Phase 9: Bulletproof it

Think through all the different possible outcomes of your referral program to make sure your strategy is bulletproof. Here are two examples of how it can all go wrong.

Upstart‌

Upstart is a site that offers people personal loans. For a short while, if you referred a friend, you and your friend would each get $150 in cash once they took out a $1,000 loan.

What happened?

People referred friends, who took out $1,000 loans. Both friends immediately claimed the referral bonus.

Then, the friend would immediately repay the loan. Without interest. This meant that Upstart made no money on the new loan. In fact, they would lose $300 ($150 for each friend).

There are hackers who spend their time trolling the internet and spreading these programs throughout their network. Don’t let them feast on you‌.

Vietnamese rats

When the French ruled Vietnam, there was a rat infestation in Hanoi. Rats everywhere. Not good.

The government came up with a bounty program that paid people a reward for each rat they killed. To claim the bounty, people would have to show a severed rat tail to prove they had killed a rat.

Suddenly, officials started seeing rats running around with no tails. It turned out that people were catching rats, cutting off their tails, and then releasing them back into the sewers. So the rats could make more rat babies. And people could cut off more rat tails.‌

Instead of improving the situation, authorities had made the problem worse.

Lesson: Be careful what you incentivize.

Run through your top ideas and make sure they can’t be easily gamed.

Phase 10: Go live

Make it happen. See how many referrals you get. Set a reminder to check if they justify the ongoing monthly costs.