Gavage: Don’t Force Feed Your Funnels

If you’re reading this, chances are your work involves dealing with funnels. Funnels are any flow with several steps where the goal is to move a person forward in some process.

Common funnels you might encounter are:

  • Marketing funnels: The classic example is customers start with awareness (of what you’re trying to sell), then move to interest (in what you’re selling), consideration, all the way through making an actual purchase.
  • Product funnels: Getting users on a website or app to take a sequence of successive steps, like signing up and onboarding.
  • Hiring funnels: You want to hire people for your company, starting from some conversation or job announcement, through to extending an offer, getting it accepted, and the person joining your company.

This article is about one of the biggest mistakes I see people making with funnels: force-feeding. It’s part of a two-part series (in this part I’ll talk about force-feeding and it’s effects, and in the second part, I’ll talk about how to avoid force-feeding).

But before we talk about funnels, let’s talk about ducks (or just skip past the next line break if you’d rather get to the funnels).

Of all the cruel and creative ways humans have found to farm / exploit, gavage is one of the most interesting. Gavage (French word) is a way of force-feeding ducks. A tube is stuck down the duck’s throat, by-passing the beak, esophagus, etc and food (corn and fat) is deposited directly into the duck’s crop.

Now, if you’re the human doing this to a duck, this is a marvelous exploitation. You can increase the weight of the duck and create wonderful delicacies like foie gras to eat or sell. It is a great short-term win for the exploiter.

We've Been Debating Foie Gras Since Ancient Times - Gastro Obscura
Apparently we’ve been doing it since ancient times (see here)

But, if you’re the duck (or a person who cares about ducks and other living things), this is atrocious. The core input mechanisms of the digestive system (beak, throat) are bypassed, and perhaps more importantly, the appetite is bypassed, too. The duck has no say in the matter. By being force-fed quantities and types of food it wouldn’t normally eat, it becomes obese, its liver swells to an unnatural proportion, and it develops a whole other slew of detrimental physical and psychological effects. I’ll spare you the rest of the gory details, but, needless to say, a duck being force-fed must be slaughtered within a certain window, otherwise it will die from the effects.

Let’s get back to sales, marketing, hiring, and product funnels. Too many people practice gavage with their funnels. They hijack and degrade their funnels by force-feeding them in unnatural ways. This can result in short-term wins, but sooner or later, the chickens come home to roost.

The chickens come home to roost - Malcolm X | Meme Generator
I’ve been trying to find a way to get Malcolm X into my writing

I’ll first talk about how this happens, then I’ll talk about the subtle effects.

Withholding Information

Let’s assume we’re tuning a product funnel that involves a subscription-based product:

  1. User sees an ad for the product
  2. User clicks the ad
  3. User lands on your landing page describing your product
  4. User signs up
  5. User pays for the subscription ($9.99 / month)
  6. (some time later) User renews their subscription

Between each of these steps, some subset of users drop off and abandons the funnel, and you can calculate a conversion rate between each step (ie what % of people who see an ad click it, or what % of people who sign up end up paying).

Let’s say that initially, your landing page mentions your pricing. Someone on your team designs an A/B test for a new landing page that doesn’t mention your pricing. The A/B test is positive! A higher percentage of people who land on your landing page are now signing up! Launch it, right? Celebrate!

Clearly, you can see the problem here. By withholding information from users, you’re kicking the can down the funnel. What will probably happen is that your down-funnel conversion rates will drop (ie more users who are now signing up, are not willing to pay, and conversion from sign-up to purchase will drop as soon as they see the price on your paywall).

Now, there’s definitely a balance here. Often, mentioning information like price too early, before you’ve built interest can deter users who might have paid had they had a better understanding of your solution. So timing is critical. But the point is, if by withholding information, you’re simplistically seeing a higher short-term conversion rate up-funnel can be misleading.

The same goes for information like salary or benefits in a hiring funnel. Too many companies and candidates keep that discussion until the end, finding out they’re miles apart after spending a large amount of time and energy.

Reducing Funnel Alignment

In general, withholding information is one way of reducing funnel alignment. Funnel alignment is probably worth its own write-up, since it’s a common mistake people make, but at a high-level, it means that at each step in your funnel, the people making it to the next step are aligned with previous steps (and ultimately, what is offered at the end of the funnel).

Funnel alignment works in two modes, usually depending on how deep a user is in the funnel. In the early stages of your funnel, lack of funnel alignment creates friction. For example, let’s say your product’s ad says: “A modern budgeting tool for the thrifty spender”, but when a user clicks on the ad, the landing pages shows a picture of a family and the text reads: “Sign up to secure your family’s future”.

Now, there are products that can help you accomplish both those things: spend less today and save for your family’s future. But a person who clicked on the first ad may not have a family. Or, they might, but now you’ve created enough ambiguity in their mind about what the product does that they go elsewhere. The two steps are not fully-aligned.

In fact, you can often get a sizable increase in conversion rates just by doing a quick scan of your early-funnel alignment. Does the ad text match the landing page message? Does the landing page message match the calls-to-action?

On the other hand, deeper down the funnel, lack of funnel alignment is less about friction and more about promise delivery. Did your product promise users that they’ll effortlessly save hundreds of dollars by using it? If not, a big chunk of your users may churn after realizing the truth. Did you promise a candidate you just hired a role that doesn’t end up materializing? They’ll probably leave at the first chance they get (and while they’re there, they probably won’t be motivated to do much of a good job).

Removing Good Friction

Growth hackers love the idea of removing “friction” from a funnel. Friction can be cognitive (you’re making someone think too much) or physical (they have to take too many steps).

In general, friction in a funnel is usually a bad thing. Too many steps or effort in your funnel, and you will lose users who might ultimately convert.

But friction can also be good. This is, again, worthy of its own write-up, but here are a few mechanisms by which “good friction” in a funnel can work:

  1. Providing information
  2. Collecting information (that might be crucial to engaging a user later)
  3. Strengthening resolve/motivation (e.g. reminding a user of certain benefits of a product)
  4. Positive energy (ie celebrating/congratulating the user)

You can often have a step in your funnel that combines several of the above mechanisms (e.g. asking a user to choose their top goals from a list collects information, strengthens resolve by focusing them on the benefits, and provides them positive energy by telling them they’ve just embarked on a path to getting those benefits. In some sense, it also provides information: the list they’re choosing from gives them an idea of the product’s supposed benefits).

So think twice before just cutting steps out of your funnel. It might seem like an easy win, but sometimes it backfires.

Lowering the Bar / Removing Qualification

In many funnels, you’re not just trying to help someone decide to choose you (or your product, company, etc), you’re also trying to choose someone. Funnels are often two-ways.

In the sales world, this is referred to as lead qualification. You have limited bandwidth to spend on leads, and so you try to determine, as early as possible, which leads are potentially going to convert to customers so you can focus your time on them.

In the hiring world, on the other hand, a good interview process performs the same function. A good interview process cuts out candidates who will not be a good fit as early as possible in the process to avoid wasting their time and your company’s time (“a good fit” could be either someone who’s not qualified for the job, or someone who’s really unlikely to accept your offer).

When pressed, sales teams (or hiring teams) will often lower the bar and perform lax qualification to push more leads (or candidates) down the funnel.

The Downsides of Force-Feeding Your Funnel

We’ve talked through a few force-feeding techniques. But what are the downsides?

Firstly, if your funnel requires resources to run, by force-feeding your funnel you’re wasting those resources. This one is obvious, and we discussed it in the qualification section above.

Secondly, you’re often fooling yourself as to the positive effects you’re experiencing. You might see better immediate conversions, but down-stream, the metrics will likely rebound and you’ll see lower conversions. And downstream metrics might be harder to measure, since they have longer feedback loops.

Often, “funnels” extend far beyond what people consider. For example, a hiring funnel isn’t just about hiring someone. It extends after a candidate joins, through their onboarding and into their employment. Hiring the wrong people will result in more attrition, but you’ll probably only find out months later. A product funnel extends after a user uses or purchases your product. They may not be retained later, or they may give you poor word-of-mouth that deters other users.

Finally, you might experience funnel corruption. Your funnel is so over-stuffed that it ceases to be a proper funnel, and causes you to make poor product decisions.

Let’s go back to that subscription product we talked about earlier. Let’s assume the company building that product engages in some drastic funnel gavage. Now, a product manager decides to launch a new feature, and, being data-driven, watches the data closely to decide on success.

That feature might be something like a new periodic email from the product with some updates. The PM expects that email to increase engagement, driving more people back to the product. She launches the A/B test. To her dismay, there’s no engagement lift. In fact, the email is driving unsubscribes and subscription cancellations! This is due to funnel corruption.

Funnels give you feedback, but that feedback is biased by the people you send down your funnel, and if your funnel is corrupt, you’ll be learning the wrong things. In this case, the people paying for the product don’t really want it, and now, every time you try to engage them, instead of getting a positive reaction, you get a negative one (because you’re reminding them about your product, which they’re paying for, but not getting value of).

Funnel corruption can destroy your product, because you end up learning the wrong things and taking the wrong actions. You start optimizing your product for the wrong people


I’ll talk about solutions in the second part of this series. Stay tuned.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: