What makes a product led loop actually loop?

What makes a product led loop actually loop?

An output of normal use has to become an input to new use, without anyone being asked to do marketing. If the loop only closes when a happy customer remembers to tell someone, it is not a loop, it is hope with a diagram around it.

Product led growth gets drawn as a circle on a slide long before anything circular exists in the product. The drawing is easy. The hard part is finding a step that people already take for their own reasons, which incidentally exposes the product to someone new.

So this piece is about the mechanics. What has to be true, where loops leak, and how to tell whether yours is compounding or just a funnel with better branding.

What is a loop, and how is it different from a funnel?

A funnel spends attention to produce customers. A loop produces its own next input, so the work of acquiring the next user is done by the previous user's normal activity. The test is whether output feeds input without new spend.

That distinction matters because the two are managed differently. Funnels improve by optimising each stage. Loops improve by shortening the cycle and reducing leakage between steps, and a loop with one broken step does not degrade gracefully, it stops.

Most B2B companies have a funnel and call it a loop. There is nothing wrong with a funnel, and pretending it is a loop leads to underinvesting in the channels that are actually producing pipeline, while waiting for compounding that was never structurally possible.

Which loops actually exist in B2B software?

Three that work repeatedly. Collaboration loops, where using the product requires involving someone else. Artefact loops, where the product creates something that gets shared outside the team. And data loops, where usage makes the product measurably better for the next user.

Collaboration loops are the strongest because the invitation is not a favour, it is a requirement. If the work genuinely cannot be finished alone, the product's own mechanics bring in a colleague. Anything where the work lives in a shared space, as with a collaborative document or design tool, has this shape naturally.

Artefact loops depend on the thing produced being worth showing. A report, a recording, a scheduling link, or a public page all travel outside the account, and each recipient is a real impression. The constraint is that the artefact has to be useful to the recipient, not just proof that your customer used a tool.

Data loops are the slowest and the most defensible, because they compound quality rather than reach. They also rarely produce acquisition on their own, so treating them as a growth loop usually disappoints.

Choosing between the three is a separate decision from making one work, and I have written about that choice directly in which loop to build first. This post assumes the choice is made and asks why the loop is not turning.

Why do most loops leak at the invite step?

Because the person being invited arrives without context. They were sent something by a colleague, they do not know what the product is, and the first screen asks them to sign up before explaining anything. That is the most expensive moment in the loop and it is usually designed last.

The fix is to make the invited person's first experience about the thing they were sent, not about your product. Let them see the document, the report, or the recording first. Signing up should be the natural next step after value has arrived, not the toll gate before it.

Second common leak is asking the inviter to do unpaid marketing. A prompt to share with your team, framed as helping you grow, converts badly because it serves you rather than them. A prompt that helps them finish their own work converts because it is genuinely useful.

Third leak is friction in the wrong place. Requiring an account to view, requiring a credit card to try, or requiring an admin to approve something all interrupt the loop precisely where momentum matters, and each of those interruptions is worth auditing separately.

What has to be true about the product for a loop to work?

Value has to arrive before the invitation makes sense, and the product has to be usable by a stranger without training. If your product needs onboarding to be useful, every loop step lands on someone who cannot get value quickly, and the loop dies at the edges.

This is why loops are harder in complex B2B tools. A product that requires configuration, integration, or permissions is not going to compound through casual exposure, no matter how good the sharing feature is. That is not a failure, it is information about which growth model fits.

The honest version of the question is whether a new person can get something useful in their first five minutes without help. If the answer is no, fix that before building loop mechanics, because everything downstream depends on it. Writing for people who will never talk to your sales team is part of the same work, which I covered in writing for buyers who never talk to sales.

How do you measure whether a loop is compounding?

Measure the cycle, not the total. How many new participants does one participant produce, how long does one turn of the loop take, and what share of new users arrived through the loop rather than through spend. Those three numbers tell you whether anything is circular.

Totals hide everything. New signups going up while the loop contributes a shrinking share means your paid and content channels are carrying growth, and the loop is decoration. That is worth knowing before you plan a quarter around it.

Cycle time is the number teams forget. A loop that produces one new participant per participant but takes six months is slower than a channel you can run this week. Shortening the cycle is often a bigger win than increasing the rate, and it is usually a product change rather than a marketing one.

What role does marketing play in a loop?

Marketing seeds it, explains it, and catches what falls out. Loops rarely start themselves, so someone has to bring the first cohort, and every invited person who does not convert immediately is a real audience your content can serve later.

The explaining part is underrated. An invited stranger arrives with a question your product cannot answer on its own, which is what your product is and whether it is for them. That is a positioning and content job, done at the exact moment of highest attention.

Then there is the recovery path. Most invited people will not convert on first contact, and the ones who do not are better prospects than a cold list because they have seen the product in use. Treating them as a nurture audience rather than a lost impression is where a lot of quiet value sits, and it connects directly to the levers that move trial to paid conversion.

When is a loop the wrong ambition?

When your product is used by one person per company, when the buying decision needs procurement, or when the artefact cannot leave the account for good reasons. Security, compliance, and confidentiality all legitimately block sharing, and no design cleverness fixes that.

In those cases the honest move is to invest in channels that work for your situation rather than forcing a mechanic that fights your product. Founder led sales, partnerships, and content built around a narrow problem all compound in their own way without pretending to be viral.

I would rather a company run two channels well than run a loop badly while telling investors it is product led. The loop story is attractive because it implies free growth, and implied free growth is the most expensive assumption in a plan.

What does a realistic first loop experiment look like?

Pick the one thing users already share, make that path excellent, and measure the cycle for a quarter. Not a referral programme, not an incentive scheme. Find the existing behaviour and remove the friction around it.

Start by looking at what leaves your product today. Exported files, shared links, screenshots pasted into chat. Each of those is a loop step happening without your help, and the fastest wins are usually in making the received version better rather than in creating a new mechanic.

Then instrument it properly so you can tell whether it worked. One path, one measurement, one quarter. If it does not move, you have learned something real about your product's shape rather than about your campaign.

What should you do next?

Write down what currently leaves your product and reaches someone who is not a user. If nothing does, you do not have a loop yet and should say so internally. If something does, spend this month making that recipient's first minute excellent.

If you want help working out whether your product can support a loop at all, and what to invest in if it cannot, that is a conversation worth having before you build the mechanics. Tell me what your users share today and I will tell you where I would look first. Let's chat.