Analytics
Free marketing budget allocator by stage and motion
By Charles Summers · Updated · Free, no signup
Short answer
This splits a monthly budget across eight channel buckets using a weighting table driven by company stage, go-to-market motion and primary goal, then reserves a slice for experiments that shrinks from about a quarter of the budget before product-market fit to under a tenth at maturity. Every line is checked against a minimum viable monthly spend, because a paid channel funded below its learning threshold buys noise, and four underfunded channels lose to two properly funded ones.
Use the marketing budget allocator
What does this tool actually do?
This splits a monthly budget across eight channel buckets using a weighting table driven by company stage, go-to-market motion and primary goal, then reserves a slice for experiments that shrinks from about a quarter of the budget before product-market fit to under a tenth at maturity. Every line is checked against a minimum viable monthly spend, because a paid channel funded below its learning threshold buys noise, and four underfunded channels lose to two properly funded ones..
It runs entirely in your browser. Nothing you type is sent to a server, no account is required, and there is no usage limit, because there is no cost per run to control.
Before product-market fit, budget buys information rather than pipeline
Allocation assumes you know which channels work, and before product-market fit you do not. Spend at that stage has a different job: producing conversations that tell you who the buyer is, what they call the problem, and which message makes them lean in. That is why the weighting here tilts pre-fit budgets towards outbound and community rather than towards paid media. Twenty conversations with the wrong people is a finding. Twenty thousand impressions is not.
The other reason to avoid heavy paid spend early is that paid channels punish small budgets in a specific, mechanical way. Automated bidding needs a volume of conversion events before it can optimise: Meta documents a learning phase of roughly 50 optimisation events per ad set per week, and Google's smart bidding guidance has historically pointed at something like 30 conversions in 30 days. A campaign funded below that never exits learning, so its results are dominated by variance and you cannot tell a bad channel from an unlucky fortnight.
This produces a rule that decides most small-budget allocations for you. Take the money you were going to give a paid channel, divide it by your expected cost per acquisition, and if the answer is below the learning threshold, do not run that channel. Consolidating into two properly funded channels beats spreading across five underfunded ones, and it is not close. The allocation this tool produces flags any line that falls below a workable monthly minimum for exactly this reason.
The same logic applies outside paid media, with different thresholds. Content below roughly one competent piece a week does not accumulate enough surface area to rank for anything competitive. An events line that cannot cover one event is a line that will be spent on something else by month three. Every channel has a floor beneath which the money does not buy a smaller version of the outcome, it buys nothing, and the floor is a property of the channel rather than of your ambition.
Product-led and sales-led budgets are not the same money relabelled
The two motions differ on where the marketing budget stops. In a sales-led business, marketing spends until a qualified meeting exists and a human takes over, so the budget concentrates on the machinery that produces meetings: outbound programmes and the data behind them, field events and dinners, category-level content that a seller can send, and the tooling that keeps a pipeline review honest. Headcount and events dominate, and the cost per opportunity is high because the deal size that justifies it is high.
In a product-led business the budget stops at signup, and the product does the selling from there. That moves money towards anything that produces qualified self-serve starts and anything that improves what happens after one: search-driven content that captures intent at scale, in-product growth surfaces, referral mechanics, and lifecycle messaging during the first fourteen days when activation is decided. Events and outbound shrink dramatically, not because they never work but because their cost per acquisition rarely reconciles with a self-serve price point.
There is an accounting trap in comparing the two that catches people constantly. Product-led companies routinely fund the work that drives their growth, onboarding flows, in-product prompts, referral systems, out of engineering and product budgets rather than marketing ones. Their reported marketing spend as a share of revenue therefore looks dramatically lower than a sales-led peer's, and the comparison is meaningless because the two are not counting the same activities. Compare total go-to-market cost against net new revenue instead, which is a definition both can be held to.
Most real companies run both motions at once, usually self-serve at the bottom and sales-assisted above some deal size. If that is you, split the budget by segment first and run this allocation twice, because a blended allocation across two motions produces a middle that serves neither: too little outbound to win the enterprise deals, too little lifecycle investment to convert the self-serve ones.
Brand and performance are different time horizons, not competing line items
Performance channels harvest demand that already exists. Brand investment changes how much demand exists in the first place, and how cheaply the harvesting channels can convert it. The most quoted evidence for a split is Binet and Field's analysis of the IPA effectiveness databank, which landed on roughly 60 percent brand and 40 percent activation as a long-run average. Worth knowing what that number is and is not: it is drawn from a body of case studies dominated by large consumer advertisers, and the authors themselves report that the optimum moves by category and by business model. Treat it as evidence that the brand share should not be zero, not as a target to copy into a B2B software plan.
The mechanism matters more than the ratio. Brand work raises the number of people who already know who you are before they start looking, which shows up later as higher branded search volume, better conversion on the same paid clicks, shorter sales cycles and less price sensitivity. Those effects arrive over quarters, and they arrive attached to activity that happened months earlier.
That lag creates the failure mode this allocation is designed to resist. Attribution systems credit the last click, so the harvesting channels always look efficient and the brand line always looks wasteful. Under budget pressure the brand line goes first, performance metrics briefly improve because there is still demand in the tank, and six to twelve months later cost per acquisition starts rising everywhere for reasons nobody can locate in the dashboard. If you cut brand, cut it deliberately with an expectation of when the cost will appear, rather than because a report that structurally cannot measure it says it is not working.
Retention marketing suffers from the same measurement problem in a harsher form. Lifecycle and onboarding work shows up as an absence, the customers who did not leave, and an absence generates no attribution row at all. That is why the retention goal in this tool moves real money into lifecycle rather than leaving it as an afterthought: for most subscription businesses past the earliest stage, a point of retention is worth more than a point of acquisition efficiency and costs less to buy.
The experiment reserve, and how to stop it being spent on something else
The reserve shrinks as the company matures for a straightforward reason: it is a purchase of information, and the value of information falls as your uncertainty falls. Before product-market fit almost everything is an experiment, so a quarter of the budget being exploratory is honest labelling rather than generosity. At scale the mix is known, the reserve stops being a search for a working channel and becomes an option on finding the next one before the current ones saturate, and under a tenth of the budget is enough to hold that option open.
The convention behind splits of this shape is the 70/20/10 model, popularised through Google's resource allocation and later through consumer marketing plans: the majority on what demonstrably works, a slice on scaling what is promising, a small slice on things that will probably fail. The specific numbers are a convention rather than a finding, but the structure encodes something true, which is that a portfolio with no failing bets is not exploring, it is just spending.
A reserve only survives if it is ring-fenced. Left unprotected it gets absorbed in week three by whichever channel is behind on its number, and that transfer always feels justified in the moment because the underperforming channel is measurable and the experiment is not yet. Protect it with three rules: the money is committed at the start of the quarter, each experiment declares its kill criterion and its minimum spend before it starts, and nothing gets moved out of the reserve to rescue a core channel.
Set expectations on hit rate too. Something like two or three winners in ten is a normal outcome for genuine experiments, and a reserve that returns eight winners was not spent on experiments, it was spent on cautious extensions of things you already knew. Size each test at whatever it takes to produce a readable result rather than dividing the reserve evenly, because three tests too small to interpret cost the same as one that answers a question, and only one of those outcomes changes what you do next quarter.
Numbers worth knowing
| Metric | Typical | What it means |
|---|---|---|
| Paid social learning phase | about 50 optimisation events per ad set per week | Meta's documented threshold. Below it, results are dominated by variance, so an underfunded channel produces noise rather than a smaller version of the outcome. |
| Paid search bidding data | roughly 30 conversions in 30 days | The figure Google's smart bidding guidance has historically pointed at. Divide a proposed channel budget by your expected cost per acquisition and check it against this before funding the line. |
| Brand versus activation split | about 60/40 in the IPA databank | Binet and Field's long-run average, drawn from a case base dominated by large consumer advertisers. Evidence that brand should not be zero, not a ratio to copy into a B2B plan. |
| Marketing as a share of revenue | roughly 7% to 11% in recent CMO surveys | Gartner's annual survey has moved within that band since 2020 and the spread by industry is far wider than the median. Useful for sanity, useless as a target. |
| Experiment hit rate | 2 to 3 winners in 10 | A reserve returning eight winners was not funding experiments. Expect most to fail and size each one to produce a readable result rather than splitting the reserve evenly. |
Mistakes that quietly cost you results
- Spreading a small budget evenly across every channel
- Below the learning threshold, paid channels return variance rather than a proportionally smaller result. Divide each proposed line by your expected cost per acquisition, and if it does not clear roughly 30 to 50 conversions a month, fold it into a channel that does.
- Running heavy paid acquisition before product-market fit
- At that stage spend should be buying information about who the buyer is and what they call the problem, which conversations produce and impressions do not. Weight towards outbound and community until the message stops changing every month.
- Cutting the brand line because attribution says it does not convert
- Last-click systems structurally cannot see work that raises demand months later. Performance metrics improve briefly after the cut while the tank drains, then acquisition costs rise everywhere with no traceable cause. Cut it deliberately or not at all.
- Copying a competitor's spend mix across a different motion
- Product-led companies fund onboarding, referral and in-product growth out of engineering budgets, so their reported marketing spend excludes the work that actually drives growth. Compare total go-to-market cost against net new revenue instead.
- Letting the experiment reserve rescue an underperforming core channel
- It always feels justified because the core channel is measurable and the experiment is not yet. Commit the reserve at the start of the quarter with a kill criterion per test, and treat any transfer out of it as a decision to stop exploring.
What does the output look like?
This is the exact output the tool produces from the example inputs. It is generated by the same code that runs when you click the button, so what you see here is what you get.
Frequently asked questions
Where do the weightings actually come from?
They are a table in the code, not a generated guess. Each of the eight buckets carries a base weight per company stage, then a motion factor and a goal factor multiply it, then everything is normalised into whatever is left after the experiment reserve. That means the same budget produces genuinely different splits across the twenty-four stage, motion and goal combinations, and the reason each line moved is printed next to it rather than left implicit.
Why does the experiment reserve get smaller as the company matures?
Because a reserve buys information, and information is worth most when uncertainty is highest. Before product-market fit almost every spend is exploratory, so about a quarter of the budget being labelled as experiment is accurate rather than generous. At scale the working mix is known and the reserve becomes an option on finding the next channel before the current ones saturate, which under a tenth of the budget is enough to fund.
What is the minimum viable spend check on each line?
It compares each allocation against the smallest monthly amount at which that channel does anything useful. For paid channels the constraint is the bidding learning phase, roughly 50 events per ad set per week on Meta and around 30 conversions in 30 days for Google smart bidding. For content it is the cost of publishing consistently, and for events it is the cost of one event. Lines below the floor are flagged so you can consolidate rather than fund noise.
Should I follow the 60/40 brand to performance split?
Not directly. It comes from Binet and Field's work on the IPA effectiveness databank, which is dominated by large consumer advertisers, and the authors are explicit that the optimum varies by category and business model. What survives translation is the mechanism: activation harvests demand that exists, brand changes how much exists, and the second effect is invisible to last-click measurement. The output shows your creation and capture split so you can see the shape rather than chase the ratio.
My company runs both a self-serve and a sales-assisted motion. What do I enter?
Split the budget by segment and run the allocation twice, once per motion, then add the results. A single blended allocation across two motions produces a middle that serves neither, with too little outbound to win the enterprise deals and too little lifecycle investment to activate the self-serve ones. If one motion is clearly dominant by revenue, run that one and treat the other as a line inside the experiment reserve.
Related free tools
- Pipeline Coverage Calculator Analytics
- A/B Test Significance Calculator Analytics
- A/B Test Sample Size Calculator Analytics
- AI B2B Headline Copywriter for hr tech Copywriting
- AI B2B Headline Copywriter for logistics Copywriting
Some links on this site are affiliate links, which means Hacking Demand may earn a commission if you buy through them at no extra cost to you. This does not influence which tools are listed. The tools on this page are free and have no affiliate relationship of any kind.