The short version
- Short-term channels buy revenue. Long-term channels buy lower future costs. Both are legitimate; they are not interchangeable.
- The common failure is spending everything short-term for years, then wondering why acquisition costs keep rising.
- The opposite failure is real too: investing in compounding channels while running out of cash before they compound.
- A defensible default for a stable business is roughly 70% short-term, 30% long-term — adjusted hard by your runway.
Every marketing channel sits somewhere on a spectrum between "pays today" and "pays later, then keeps paying". Paid search is at one end: switch it on, get customers, switch it off, they stop. SEO, content and brand sit at the other: months of nothing, then an asset that produces at declining marginal cost.
The reason businesses get this wrong is that the two ends are measured on different timescales but argued about in the same meeting. Short-term channels always look better in a quarterly review, because that is the timescale they operate on. A long-term channel evaluated quarterly will lose every time, right up until the moment it would have started working.
The short answer
Fit for your workflow
Let runway decide. Short runway buys short-term channels; long runway buys compounding ones.
If you have under six months of cash, compounding channels are a luxury you cannot afford — spend on what converts this month. If you have a stable business with predictable revenue, spending nothing on long-term channels means your acquisition costs will keep rising and you will have no defence against it. Most stable businesses are structurally under-invested in the compounding end.
- Under 6 months runway
- Almost entirely short-term. Survive first.
- 12+ months, growing
- ~70/30 short/long. Start the compounding clock now.
- Profitable and stable
- ~60/40. You can afford patience and your competitors probably cannot.
- Acquisition costs rising year on year
- Shift toward long-term. That trend is the symptom of under-investing in it.
The spectrum
The "stops when you stop" column is the one to reason about when budgets tighten. Cutting paid spend removes revenue this week. Cutting SEO removes nothing this week, which is exactly why it gets cut first and why the consequences arrive two quarters later, uncorrelated in time with the decision that caused them.
The trap in each direction
All short-term
The business grows on paid acquisition, cost per acquisition creeps up each year as competitors bid, and margin compresses. There is no owned audience and no organic presence to fall back on, so any increase in auction pressure goes straight to the bottom line. This is an extremely common shape and it is comfortable right up until it is not.
All long-term
The business invests in content, brand and SEO, produces genuinely good work, and runs out of money four months before any of it compounds. This failure is rarer but not rare, and it is usually the result of treating "long-term is better" as a principle rather than a function of runway.
Making the case internally
Long-term channels lose budget arguments because they are evaluated on a short-term reporting cycle. The fix is to agree the measurement basis before the spend starts, not when the first quarterly review goes badly. Commit to leading indicators — indexed pages, impressions, query breadth, changes shipped — and state explicitly that revenue is not expected for two to three quarters.
It also helps to anchor the conversation in something neutral. Google’s own starter guide describes organic search as a long-term investment rather than a campaign, which is a more persuasive citation in a finance meeting than an agency deck. Pair that with your own trend in cost per organic visit and the argument stops being a matter of faith — see SEO cost vs ROI.
A simple annual check
Compare your blended cost per acquisition this year against last year. Rising means you are under-invested in compounding channels and paying more each year for the same customers. Falling means something is compounding — find out what and fund it more.