Customer referral programs are essential to growth. In fact, I think they are one of the most underrated ways to unlock the positive network effects needed to achieve product/market fit.

Network effects relate to the phenomenon that increased usage by any user increases the value of the product or service for other users. I think about this a lot, particularly in the context of my new venture. What’s often missed about network effects is that they generate a positive feedback loop and as a consequence, growth.

Network effects 101

There are five broad types of network effects. The tactics that companies employ to generate network effects in the hope of tapping significant growth are often seen as interesting conveniences to their customers and users. The ‘Sign up to be a driver’ message in the Uber app or the ‘Import your address book’ feature on LinkedIn, are the tactics that directly or indirectly increase the value of the service to others. Airbnb, Facebook and Atlassian (and many other software companies) also employ now iconic tactics to drive network effects.

The trick that founders often miss is that while the concept of network effects is relatively easy to grasp, they think they can emulate Uber, LinkedIn or Instagram’s tactics to generate similar growth trajectories. That almost never works for three reasons.

First, time moves on. There was a time when many people wouldn’t have thought twice about clicking ‘Import’ and allowed LinkedIn to ingest their entire address book in exchange for a broader reaching and more comprehensive online professional network. Today, providing any kind of information to a third party site has people pausing to consider the implications.

Second, the tactic doesn’t match the context. Just because LinkedIn grew thanks to people importing their address books doesn’t mean that function will make sense (or be useful) to your customers and users. Still, founders look at the iconic companies and think if they can do it, we can do it too!

Third, early adopters are more selective. The world has now experienced a decade of mobile consumer apps and services. Companies like LinkedIn and Facebook were early to the party and able to woo early adopters and leverage their interest using tactics that drove network effects and growth. The world has changed and early adopters are more familiar (and resistant) to these tactics.

Network effects come from two things

When I design network effects into business modes, I reflect on how Reid Hoffman describes them as coming from two things: leveraging existing networks and virality.

Before I talk about leveraging existing networks, I think it’s important to break down a common misconception about virality. When you hear that a piece of online content or a product has ‘gone viral’, people can sometimes think that word-of-mouth was responsible for its rapid accession into people’s consciousness. As a result, they start deconstructing what made the content or products word-of-mouth so potent.

As Alexander Jarvis says, word-of-mouth is just that, a human telling another human about something of interest. Whereas virality is a product phenomenon that is engineered by thousands of little optimisations requiring product design, a growth hacker like mentality and a keen understanding of human behaviour.

It’s important to understand this distinction. Founders should also be clear on the reality that vitality is inherently easier to achieve with free and freemium business models and as a result, better suited to software products.

Referral programs and network effects

Referral programs leverage existing networks and as a consequence are one of the most potent drivers of network effects. Unlike shares and likes, referral programs can build on a person’s interest after they have signed up, downloaded or paid for a product or service. And when done well, referral programs accelerate growth by reducing customer acquisition costs (and therefore increasing margins that can be reinvested into growth), exposing new customer segments, creating competitive moats and increasing k-factor (more on that later).

Many founders talk a good game when it comes to network effects. Unfortunately, there is often a wide gap between talk and execution. I look at the quality and attention paid to a venture’s customer referral program to see how wide that gap actually is. For the most part referral programs, if one exists, are bolted onto the customer experience and that’s not a good thing.

My theory is that founders pay little attention to referral programs for two reasons.

First, they are too busy building their product and are overly confident in their ability to market and distribute it. This is often characterised by founders thinking, ‘you can’t sell if you don’t have a product’ or that they’ll get to customer acquisition later and believe paid advertising will solve their distribution challenge.

Second, they think of customer referral programs (if at all) as just another marketing channel and like building the product, they think it can add it as a feature.

I find this mindset ironic given how many startups fail due to poor product distribution. In any case, these responses are symptomatic of founders not properly understanding their customers or how to drive growth.

Why referral programs fail

I can think of seven reasons why referral programs fail, all of which are linked to a lack of personalisation.

  1. One-sided offers: You’re asked to share a product or service with your friends and network and while referee may get a financial incentive (e.g. a discount) or a non-financial inventive (e.g. early or free access) and the company may get a new customer, you get nothing
  2. Links look and feel like SPAM: In this case, the link you are asked to share is a long and random looking string. It’s 2018. Why would anyone click on a link that looks like SPAM? I think the Trello referral link is the best in class. It’s personalised and notwithstanding the quality of the product, it encourages me to share it:
  3. Nuclear is the only share option: Social only share options in referral programs are the best way to kill sharing and are the best way to demonstrate that your referral program is bolted on. And there are some that reinforce this over time. I recently purchased a direct-to-consumer biotechnology product, shared the referral code on Facebook and now each month, I receive an email from the referral program provider, not the biotechnology company, that no one wants to take up the offer. Really?   
  4. A clueless offer of value: The classic example of this reason why referral products fail is that when the incentive is in a value or language that the company understands but which makes absolutely no sense to anyone else
  5. One ask too many: This tends to happen when physical product companies ask new customers to refer them to friends before the customer has received any value. Customers have searched for the product, invested in setting up an account, paid for the product and then been asked to commit an act of referral. All while not yet having received value. Timing is everything.
  6. Locked-in incentives: This is more a mindset issue where founders set and forget the referral incentives (or they think the incentives can’t change). In the first scenario, the referral program was considered a feature to buy (or build) and then ship as part of a product release. The incentive is fixed and nothing changes. The scenario at the other end of the spectrum is that the incentive is so successful that founders are reluctant to adjust it, even if they are haemorrhaging cash or cannibalising their customers or users in the short-term. In both cases, the cardinal rule of product development applies; always be learning and course correct based on data
  7. It is actually bolted on: In this scenario, the referral program is so obviously bolted on that you are redirected to a page which could not be a bigger departure from the page you were just on.   

While these points are important, they are only part of what you need to consider before designing a referral program.

Know your k-factor

While ‘k’ as the designation for k-Factor remains a mystery to me, its importance is not.

K-factor is borrowed from virology where a virus is said to be in a steady state, not growing or declining when k equals one. When a k-factor is greater than one (which can be 1.00000001) indicates exponential growth while a k-factor less than one suggests an exponential decline.

This biological framework helps when it comes to marketing and referral programs. The math to estimate k-factor is relatively simple:

k = i x c where i is the number of invites sent by each customer and c is the percent conversation of each invite

If each new customer invites 10 friends (i = 10) and if two invitees convert to new customers (c = 0.2), k-factor = 2.

While there are additional factors that help hone the k-factor, like market size and churn, focusing on the basics when designing a referral plan is a good start.

Build on an advantage

I mentioned earlier that one of the reasons that founders pay little attention to referral programs is they think they are but one of many marketing channels. It is only one marketing channel but I suspect that that one to one relationship is based on a perception that a referral program can only ask one question. In other words, the link you ask users and customers to share can only ask them to sign up in exchange for an incentive.

Well, that’s not true.

Depending on your company and business model, you can consider a multifaceted referral program which can include:

It all comes down to imaging, in vivid detail, the best experience you want your target customer or user to have. And then going two steps further, imagining how that experience makes them feel and then capitalising on that feeling by encouraging them to be your advocate.   

One last thing…

I love a well-designed customer referral plan because it tells me a lot about how a company is thinking about its customers.

I also think it’s important to be clear on what a referral program is and what it’s not. Compelling referral programs help people invested in a product or service to introduce it to others again and again. However, referral programs will not help a company grow if their product fails to meet a need.

The main point I’d like you to leave with is that a referral program should go live the moment a product is launched. It can start with founders manually emailing a personalised referral code from a spreadsheet to new users and customers. Design it to avoid the seven factors that would otherwise make it fail.

When executed well and continuously improved, a referral program can help reveal new customer segments and provide clues as to why a product isn’t growing as organically or as quickly as you had initially hoped.