Network Effects Intensity Mapping
Map the network's shape before you claim network effects give you power
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 90%
Helmer's objection is not to network effects but to their casual assertion. Because networks are complex, saying a business has them gets you a fraction of a percent of the way to understanding what is going on. The method is to map the network explicitly: who connects to whom, whether the sides are symmetrical, whether each participant values all others equally, and how marginal value scales. The decisive variable is the shape of the value curve. If value rises steeply and never flattens, the largest network is unassailable. If value flattens at a level a second competitor can reach, and the market is big enough to support more than one player in that flat region, the network effect produces real customer utility and no power. Density economics matter for the same reason: a regionally bounded network effect must be won market by market and does not transfer.
Origin
Extracted from Acquired, where Helmer applies the mapping method live to Uber and explains why utility that transformed his own travel does not automatically translate into a durable profit pool.
Core principles
- 01Asserting network effects tells you almost nothing; intensity is the whole question
- 02Networks are rarely homogeneous, linear or symmetrical
- 03Value that flattens with scale leaves room for a second winner
- 04Utility to users and monetisable power are different questions
- 05Geographic or density-bounded networks do not compound across markets
How to run it
- 1
Draw the map
Write down every participant class and the specific value each derives from each other class. For a marketplace this means naming what a driver gains from more passengers and what a passenger gains from more drivers.
Pro tip Do this on paper before touching a spreadsheet; the asymmetries usually appear immediately.
- 2
Test for heterogeneity and asymmetry
Ask whether all participants are interchangeable to the other side. Most networks have participants who matter far more than others, and sides whose need for the other is unequal.
Pro tip Ask which specific participants a new user actually cares about, not how many.
Watch out Treating the network as homogeneous will overstate the effect badly.
- 3
Plot marginal value against scale
Model how much better the product gets with each additional participant. The relationship is usually nonlinear and the interesting question is where it stops improving.
- 4
Locate the flattening point
Find the scale at which additional participants stop meaningfully improving the experience. Beyond that point, being bigger is no longer being better.
Pro tip For local networks like ride-hailing, the flattening point is often a wait-time threshold measured in minutes.
- 5
Ask how many players fit past the flattening point
If a similarly sized competitor can reach the flat region and the market is large enough to sustain both, the network effect confers no advantage. This is Helmer's first-order reason for doubting Uber's network-effect power.
Watch out Two players in the flat region typically means competition on price, not power.
- 6
Separate utility from monetisation
Map the desire to join the network to your actual ability to monetise it, which depends on engagement and willingness to pay. Helmer insists everything maps back to value or it is not strategy.
Pro tip Write the mapping function explicitly: participants, to engagement, to revenue, to margin.
In the wild
Helmer describes the benefit vividly: after board meetings in Vancouver, where no ride-hailing operates, a taxi is called and he has no idea whether it is coming or where it is, so he gives up and hails a cab on a corner. The utility is enormous. But the network's density economics are regional and nonlinear: past a certain driver density in a city the experience stops improving, a similarly sized competitor can reach that same density, and the market is large enough for more than one. First order, he would say Uber does not have power from network effects.
→ Powerful customer benefit, minimal network-derived barrier, because the value curve flattens where a second player can reach.
A startup builds a hiring marketplace for a specialised trade. Mapping the network shows employers care about a specific subset of certified candidates, not headcount, and candidates care only about employers in their region. Marginal value flattens after a few hundred certified profiles per metro. That analysis reframes the strategy: instead of chasing raw sign-ups, the team competes on credentialing depth and per-metro density, and pairs the effect with switching costs embedded in employer workflows.
→ The map redirects effort from a vanity growth metric to the variable that actually creates a barrier.
Common mistakes
Asserting the effect instead of mapping it
Saying a business has network effects is a label, not an analysis. Helmer's point is that the assertion gets you a rounding error's worth of the way to understanding what is actually going on.
Assuming linear and symmetrical value
Networks are usually neither. Assuming every added participant is equally valuable to every other one overstates the barrier and hides the flattening point where competition returns.
Mistaking user delight for defensibility
A product can transform the customer experience and still be unable to monetise the advantage, because a rival can deliver the same experience at the same scale in the same market.
Is it for you?
Best for
Evaluating marketplaces, platforms and social products where the network-effects claim is doing heavy lifting in the thesis.
Not ideal for
Single-sided products with no participant-to-participant value, where the relevant powers are scale or switching costs instead.
From the transcript
“it's the intensity of the network effect that really matters”
“they're not homogeneous, they're not necessarily linear, they're not symmetrical”
“if the market's large enough for more than one company to exist in that flat spot, then it doesn't help you”
From the episode
7 Powers with Hamilton Helmer