eCommerce Growth Guide
Most growth advice is really channel advice. This guide is about the business the channels sit inside: the five terms you can actually move, what matters at your stage rather than at somebody else’s, why cash and inventory kill more growing stores than competition does, and the operating cadence that turns all of it into decisions.
CHAPTERS
8 chapters
READING TIME
37 min read
LAST UPDATED
19 September 2026
LEVEL
Intermediate
Growth is five numbers multiplied together. Most teams only ever work on the first one.
EXECUTIVE SUMMARY
What this guide argues, in five points
An eCommerce business grows when one of five numbers moves. Almost all the attention, budget and anxiety goes to the first of them, which is also the most expensive and the least durable. This guide is about the other four, and about the constraints — cash, stock, attention — that decide how fast any of them can move.
- Growth is traffic multiplied by conversion, order value, purchase frequency and margin. Four of those five improve the business permanently; one of them you rent.
- What matters depends on your stage. The work that takes a store from nothing to its first million is not the work that takes it from five to fifteen.
- Order value is the least-worked term and often the fastest. It requires no new customers and no additional media spend.
- More stores are killed by cash and stock than by competition. Growth consumes working capital before it produces profit, and the gap is where businesses fail.
- A monthly cadence with a small number of honest numbers beats a dashboard with thirty widgets and no decisions attached to any of them.
The eight chapters below are the detail behind those five points, in the order we work through them when we are brought in to grow a business rather than a channel.
- Contents
Eight chapters
Read it in order the first time. After that, the chapter index is the useful part.
01
Five numbers multiplied together. Four of them improve the business permanently; one of them you rent.
5 min
02
Four stages, four different constraints. A great recommendation at the wrong stage is indistinguishable from bad advice.
5 min
03
The least-worked term, and the one where every extra pound arrives at close to full margin.
5 min
04
Growth consumes working capital before it produces profit. A profitable business can run out of money, and many do.
6 min
05
Not a prediction and not a target — a shared assumption, written down, that makes being wrong visible early.
5 min
06
Four rhythms at four altitudes, each ending in a decision. This is what turns the previous chapters into growth.
4 min
07
Not in-house versus agency — which capabilities are the business, and which are better rented.
4 min
08
Four situations where pursuing growth is the most expensive decision available, and what to optimise instead.
3 min
- Before you start
Who this guide is for
Written for the people who own the whole number, not one channel of it.
Founders and owners
You want to know which term of the equation is movable at your stage, and how much cash the answer will require.
Heads of eCommerce and general managers
You are being asked for a growth plan and want one that survives a conversation with finance.
Finance and operations leads
Chapters four and five are your half, and they decide how fast anything in the others is allowed to move.
NOT WRITTEN FOR
People looking for channel tactics. Those live in our marketing guide. This one is about the business those channels sit inside — the equation, the stage, the cash and the cadence.
CHAPTER 01
5 min
The growth equation, and which term you can move
Five numbers multiplied together. Four of them improve the business permanently; one of them you rent.
Revenue in an eCommerce business is five numbers multiplied together. Writing them out as a product rather than as a list changes how a growth plan gets built, because a ten per cent improvement in any of the five produces the same revenue effect — and four of them cost far less than the first.
Sessions
Traffic
Rented. Stops the day you stop paying, unless it is organic.
Conversion rate
How many buy
Permanent. Improves every channel at once.
Average order value
How much per order
Permanent. Needs no new customers.
Purchase frequency
How often they return
Permanent. Raises what you can pay for acquisition.
Contribution margin
What you keep
Permanent. Decides whether any of the above is worth having.
The same revenue outcome, at wildly different costs. Most growth plans work on row one exclusively.
KEY TAKEAWAY
Compounding comes from the multiplication. Improving three terms by ten per cent each is a thirty-three per cent revenue increase, and each improvement is smaller and more achievable than a single thirty-three per cent traffic increase.
Working out which term is movable
1
Write down all five for the last twelve months. Monthly, one sheet. Most businesses have never seen them side by side.
2
Ask what a ten per cent improvement in each would be worth in contribution, not revenue.
3
Ask what a ten per cent improvement in each would cost in money, time and management attention.
4
Rank by value divided by cost. Traffic almost never wins this ranking, which is why it is startling the first time a team does it.
The output is a single sentence: which term you are working on this quarter and why. Everything in the rest of this guide assumes you have written that sentence, because a plan that improves all five simultaneously is not a plan.
Our twelve-point growth assessment turns this into a scored diagnosis, with a threshold attached to each check.
CHAPTER 02
5 min
What actually matters at each stage
Four stages, four different constraints. A great recommendation at the wrong stage is indistinguishable from bad advice.
Most growth advice is written at one stage and read at another, which is why so much of it feels both true and useless. The work that takes a store from nothing to its first million is not the work that takes it from five million to fifteen, and applying the second set early is a reliable way to run out of money.
The constraint at each stage is different. So is the right answer to almost every question in this guide.
Four stages, four different jobs
Stage
The real constraint
What actually matters
Finding it
Proof that anybody wants this
One product that sells, one channel that works, honest unit economics
Making it repeatable
Doing it again on purpose
A second channel, repeat purchase, stock that does not run out
Scaling it
The business outgrowing its systems
Forecasting, cash, hiring, and the operating cadence in chapter six
Defending it
Margin erosion and competition
Brand, retention, own-channel share, and product development
The most common mistake is running stage-three tactics in a stage-one business: hiring, tooling and channel breadth bought before the first thing is proven.
KEY TAKEAWAY
Name your stage before choosing a tactic. A great stage-three recommendation applied to a stage-one business is indistinguishable from bad advice.
The transitions, which is where it breaks
- Finding it to repeatable: the founder stops being the system. Things get written down, and the first person is hired for the part that is already working.
- Repeatable to scaling: forecasting becomes real. Stock, cash and lead times start deciding what is possible before marketing does.
- Scaling to defending: growth rate slows and margin becomes the number that matters. Teams that keep optimising for growth rate here usually buy it with margin they cannot spare.
Each transition looks like a plateau from the inside. It usually is not; it is a business that has run out of the thing that worked at the previous stage and has not yet built the thing that works at the next one.
CHAPTER 03
5 min
Order value and pricing
The least-worked term, and the one where every extra pound arrives at close to full margin.
Average order value is the least-worked term in the equation and frequently the fastest to move. It requires no new customers, no additional media spend and no change to the funnel — and every pound added to it flows through at close to full contribution margin, which is not true of a pound of new revenue from traffic.
Five ways it moves, in rough order of effort
Lever
What it changes
Watch out for
Free-delivery threshold
Basket size, immediately
Setting it below your median order value, which changes nothing
Bundles and multipacks
Units per order and fulfilment economics
Bundling your best seller with something nobody wants
Genuinely relevant cross-sell
Attach rate at the decision point
Generic recommendations, which train people to ignore the module
Good, better, best ranges
Mix, by giving people something to trade up to
Only offering one option, which makes price the whole decision
A price increase
Margin and order value at once
Doing it silently across the whole range rather than testing on a segment
The first row is a setting. The last is a decision. Most stores have never deliberately made either.
KEY TAKEAWAY
Set the free-delivery threshold above your current median order value, not below it. Below it, you are paying for delivery on orders that would have happened anyway.
Pricing, which nobody wants to touch
A modest price increase is the single highest-leverage action available to most eCommerce businesses, because it lands entirely in contribution. A five per cent increase on a thirty per cent contribution product raises contribution by roughly a sixth, before any volume effect. The volume effect is real and worth testing — but it is usually smaller than the fear of it.
- Test on a segment or a subset of the range first, and watch conversion rather than opinion.
- Change the offer at the same time where you can. A better guarantee, faster delivery or a larger size makes an increase legible rather than opportunistic.
- Do not discount to defend a price rise. Promotional discounting is a price decrease with extra steps, and it trains customers to wait.
- Watch contribution, not revenue, for a full purchase cycle afterwards.
CHAPTER 04
6 min
Cash, inventory and the growth that kills
Growth consumes working capital before it produces profit. A profitable business can run out of money, and many do.
More growing eCommerce businesses fail on cash than on competition. Growth consumes working capital before it produces profit: stock is paid for before it sells, marketing is paid for before the order arrives, and the faster you grow the wider that gap becomes. A profitable business can run out of money, and many do.
Stock is cash in a box. How long it sits there is a growth constraint, not a warehouse detail.
The cycle that decides how fast you can grow
Pay the supplier
Day 0
Cash leaves
Stock arrives
Day 30 to 90
Lead time, and it is rarely the quoted one
Stock sells
Day 60 to 180
Depends entirely on turn rate
Payout received
Day 65 to 195
Marketplace and gateway settlement terms add days
The distance between the first and last rows is how much cash growth requires. Shortening it is a growth lever.
KEY TAKEAWAY
Inventory turn is a growth metric. A business turning stock six times a year can fund roughly twice the growth of one turning it three times, on identical margin and identical capital.
What actually shortens the cycle
- Negotiate supplier terms before negotiating price. Thirty days of credit is frequently worth more to a growing business than two per cent off the unit cost.
- Order more often in smaller quantities where the unit economics allow. The bulk discount that ties up six months of cash is rarely the bargain it looks like.
- Kill the slow-moving tail. Products that turn once a year are cash you have already spent, sitting still. Clear them and redeploy.
- Forecast stock against a marketing plan, not against last year. A campaign that sells out in week two costs more than it earns.
- Know your settlement terms on every channel. Marketplaces and gateways hold money for longer than founders expect.
The practical test: work out how much cash a twenty per cent increase in volume would require, before it produces any profit. If the answer is more than you have, twenty per cent growth is not available this quarter regardless of how good the marketing plan is — and that is a useful thing to know in advance rather than in month three.
CHAPTER 05
5 min
Forecasting and planning
Not a prediction and not a target — a shared assumption, written down, that makes being wrong visible early.
A forecast is not a prediction and it is not a target. It is a shared assumption about what will happen, written down so that everybody buying stock, booking media and hiring people is working from the same set of numbers — and so that being wrong is visible early enough to act on.
Build it from the five terms
1
Start from sessions by channel, with the organic line grown conservatively and the paid line tied to a budget you have actually committed.
2
Apply conversion rate by channel and device, not a blended figure. The mix change is usually where forecasts go wrong.
3
Apply order value and expected mix, including any promotional periods that shift it.
4
Add returning customer revenue separately, built from cohort repeat rates rather than a percentage guess.
5
Convert to contribution, then to cash with the timing from chapter four. Revenue forecasts that never become cash forecasts are how stores get surprised.
KEY TAKEAWAY
Forecast contribution and cash, not just revenue. A revenue forecast that comes true while cash runs out is a forecast that was measuring the wrong thing.
Three rules that keep it useful
- Re-forecast monthly, keep the original. The gap between the two is the most useful management information in the business.
- Write the assumptions next to the numbers. A forecast without assumptions cannot be debugged when it is wrong, only argued about.
- Build a downside case and decide the trigger. What you would stop doing at minus twenty per cent, agreed in advance, is worth more than a more accurate base case.
Seasonality deserves its own note: most eCommerce businesses have a shape to their year, and a plan that spends evenly across it is wrong twice. Build the forecast on your own last two years rather than on a category benchmark, and fund the peak from the trough rather than from optimism.
CHAPTER 06
4 min
The operating cadence
Four rhythms at four altitudes, each ending in a decision. This is what turns the previous chapters into growth.
Everything in the previous five chapters only produces growth if somebody looks at it on a schedule and changes something as a result. The businesses that compound are rarely the ones with better ideas; they are the ones with a cadence that turns numbers into decisions before the quarter is over.
A meeting that ends without a decision is a status update. Growth comes from the decisions.
A cadence that works at most sizes
Rhythm
What is reviewed
What comes out of it
Weekly, 30 minutes
Sales against forecast, stock risks, anything broken
Immediate fixes and pacing changes
Monthly, 90 minutes
The five terms, contribution, cash, cohort repeat rate
One change to the plan, owned and dated
Quarterly, half a day
Which term you are working on, and whether that is still right
The next quarter’s single priority
Annually
Stage, range, pricing, partners, capital
The shape of the year and the budget behind it
Four rhythms, each with a different altitude. Collapsing them into one weekly meeting produces a lot of talking about last week.
KEY TAKEAWAY
Every review should end with something to stop doing. A cadence that only ever adds work is a cadence that will be abandoned within two quarters.
The numbers that belong in the monthly
- The five terms of the equation, with last month and last year beside them.
- Blended contribution and blended CAC, reconciled to the accounts rather than summed from platforms.
- New versus returning revenue, separated, so growth is distinguishable from harvesting.
- Cash position and inventory turn, which together decide how fast anything else is allowed to move.
- Forecast versus actual, with the variance explained in one sentence per line.
Five things, one page. If the monthly pack runs to thirty slides, the discussion will be about the pack rather than about the business, and the decision that was supposed to come out of it will be deferred to next month.
CHAPTER 07
4 min
Team, agencies and what to keep in-house
Not in-house versus agency — which capabilities are the business, and which are better rented.
At some point growth stops being a question of what to do and becomes a question of who does it. The decision is not really in-house versus agency; it is which capabilities are so central to your business that they must live inside it, and which are better rented from people who do them every day.
A test for what belongs in-house
- Does it touch the customer relationship or the product? Merchandising, service, brand and range decisions belong inside. They are the business.
- Does it require deep knowledge of your catalogue? Keep it close, or accept a long ramp for whoever you bring in.
- Is it a specialism with a steep, changing learning curve? Paid media, technical SEO and development are the usual candidates for renting.
- Can you keep one person busy and good at it? A half-time specialist is usually a specialist who falls behind, and the work looks fine until it suddenly does not.
KEY TAKEAWAY
Never outsource the part you cannot evaluate. If nobody internally can tell whether the work is good, no agency relationship will save you — and the first honest conversation will come a year late.
Hiring in the right order
1
Hire for the thing that is already working and constrained by the founder’s time. That is the hire that pays back fastest.
2
Then hire operations, because growth breaks fulfilment, stock and service before it breaks marketing.
3
Then hire the specialism you use most, and only once there is enough of it to fill a role properly.
4
Keep an internal owner for anything you rent. Not to do the work, but to judge it and to hold the numbers.
The honest version of this chapter from an agency: the engagements that work are the ones where somebody on the client side owns the number and can argue with us about it. The ones that fail are the ones where the work was handed over entirely and nobody inside the business could tell whether it was any good.
Our how-we-work page sets out what we take on, what stays with you and what the first ninety days look like.
CHAPTER 08
3 min
When growth is the wrong goal
Four situations where pursuing growth is the most expensive decision available, and what to optimise instead.
Not every business should grow as fast as it can, and an agency saying so is unusual enough to be worth saying plainly. Growth is a means. There are situations where pursuing it is the most expensive decision available, and recognising them is part of running the business well.
Four situations where growth is the wrong goal
Situation
What growth does
What to do instead
Contribution is negative or marginal
Multiplies the loss
Fix pricing, mix or cost base first
Cash is the binding constraint
Accelerates the shortfall
Shorten the cash cycle before adding volume
Service is already failing
Converts new customers into detractors
Fix fulfilment and support; they are retention
The owner wants a business, not an exit
Buys scale with margin and freedom
Optimise for profit per hour and durability
The last row is the one nobody writes about. A smaller, more profitable, less fragile business is a legitimate objective.
KEY TAKEAWAY
Growth that costs more than it returns is not growth; it is a subsidy you are paying to your customers and your platforms. Contribution is the test, and it is worth applying before the plan rather than after it.
What to optimise instead
- Profit per order, which usually responds to range and pricing faster than to volume.
- Fragility, which means channel concentration, supplier concentration and key-person risk. All three are quietly more dangerous than slow growth.
- Owner time, which is a real constraint and rarely appears in any plan.
- Optionality, which is what a cash buffer and low fixed costs actually buy you when a channel or a supplier changes without warning.
EXPERT VIEW
“The most useful thing we have told some clients is that the growth plan they asked for would make their business worse. A smaller business that keeps its margin and its owner is a better outcome than a larger one that keeps neither.”
Dazzle Commerce · Growth strategy team
- Worked examples
Three decisions, before and after
The three arguments that come up on almost every growth plan we are asked to review.
WHICH LEVER
Buying traffic to fix an order value problem
Before
Media budget raised 40% with average order value flat at £46
After
Budget held; delivery threshold moved above median, two bundles launched
Order value moved 11% in a quarter at almost no cost, and every pound of it arrived at full contribution. The media increase would have bought the same revenue at a third of the margin.
CASH
A growth plan the balance sheet could not fund
Before
Plan for 45% growth on 90-day supplier terms and a 3x stock turn
After
Plan rebuilt at 20%, with supplier terms renegotiated and the slow tail cleared
The original plan needed more working capital than the business had. Twenty per cent that is funded beats forty-five per cent that runs out of money in month five.
OBJECTIVE
Growth as the default goal
Before
Target set at 30% revenue growth, contribution margin unexamined
After
Target set on contribution, with revenue growth allowed to land where it lands
Revenue growth was being bought with discounting and paid media at negative contribution. Changing the target changed the decisions underneath it within one quarter.
Anonymised from businesses we work with. The shapes recur; the numbers will not match yours exactly.
- Take these with you
Three checklists
Written to be worked through rather than read. Nothing here needs a tool you do not already have.
The equation, once a quarter
Ninety minutes with twelve months of data.
- All five terms written out monthly for the last twelve months
- Value of a 10% improvement in each, in contribution
- Cost of a 10% improvement in each, in money and attention
- Terms ranked by value divided by cost
- This quarter’s single priority written in one sentence
- Stage named, and tactics checked against it
Cash and stock, monthly
The constraint that decides how fast anything moves.
- Cash conversion cycle measured end to end
- Inventory turn calculated, by product group not blended
- Slow-moving tail identified and a clearance decision made
- Supplier terms reviewed, credit negotiated before price
- Settlement terms known for every channel
- Working capital required by the growth plan calculated
The monthly pack
One page, five things, one decision.
- The five terms with last month and last year beside them
- Blended contribution and blended CAC, reconciled to the accounts
- New versus returning revenue, separated
- Cash position and inventory turn
- Forecast versus actual, variance explained in one sentence per line
- One thing to stop doing, agreed and dated
Copy these into your own tracker. There is no download form and no email wall.
- Frameworks
Two frameworks worth keeping
The two decisions that get reargued every quarter, written down so they stop being reargued.
FRAMEWORK ONE
The lever ranking
Every growth argument is really an argument about which term to work on. Rank them the same way every quarter: what a ten per cent improvement is worth in contribution, against what it costs to get. Traffic rarely wins, which is the point of doing it.
Term
Typical cost to move 10%
Durability
Traffic
High — media spend, or two quarters of organic work
Rented, unless organic
Conversion rate
Medium — template and policy work
Permanent, and multiplies every channel
Order value
Low — thresholds, bundles, range, price
Permanent, at near-full margin
Purchase frequency
Medium — flows, product cadence, service
Permanent, and raises affordable CAC
Contribution margin
Low to medium — pricing, mix, cost base
Permanent, and decides everything else
Costs vary by business, which is why you rank your own rather than adopting this one. The ordering is what usually surprises people.
Pick one term per quarter. A plan that improves all five at once is a list of intentions.
FRAMEWORK TWO
The growth priority order
When a business needs everything doing, the order matters more than the list. Each step assumes the one above it has cleared, because a growth plan at step six is unfundable while step two is wrong.
1
Contribution. Know it per product group. Growth on negative contribution multiplies a loss.
2
Cash. Measure the cycle and the working capital a plan would consume. This sets the speed limit.
3
Stage. Name it, so the tactics chosen belong to the business you actually have.
4
The lever. Rank the five terms and pick one for the quarter.
5
The plan. Built from the five terms, converted to contribution and then to cash.
6
The cadence. Weekly, monthly, quarterly, annual — each ending in a decision.
7
The team. Hire for what is already working and constrained; rent the specialisms.
If any step produces the answer that growth is the wrong goal right now, that is a legitimate result of running the order properly.
- In summary
Key takeaways
Eight sentences you could hand to somebody who is not going to read the guide.
1
Growth is five numbers multiplied together. Four of them improve the business permanently; traffic is the one you rent.
2
Rank the five terms by value divided by cost every quarter, and pick one. Traffic rarely wins that ranking.
3
Name your stage before choosing a tactic. Stage-three tactics in a stage-one business are how money runs out.
4
Order value is the least-worked term, and every pound added to it arrives at close to full contribution margin.
5
Set the free-delivery threshold above your median order value, not below it.
6
Growth consumes working capital before it produces profit. Inventory turn and supplier terms are growth levers, not admin.
7
Forecast contribution and cash, not just revenue, and keep the original forecast next to the re-forecast.
8
Every review should end with one decision and one thing to stop. A cadence that only adds work gets abandoned.
If you want a single first action: write all five terms out monthly for the last twelve months on one sheet. Most businesses have never seen them side by side.
- Go deeper
Where this work sits in the practice
Three services that pick up the chapters above.
The twelve-point diagnosis that names which term of the equation is actually your constraint.
Catalogue, stock, fulfilment and service — the operational half that chapter four makes the speed limit.
The channel work, run against contribution once the equation has told you which term to move.
- Keep reading
Related guides
The rest of the guide library, all published in full — plus the revenue, acquisition cost and lifetime value calculators behind the growth equation.
Complete guide · 37 min
The channel half: which channel does which job, and how to measure it when attribution is broken.
Complete guide · 37 min
The second term of the equation in full, including how to test it at realistic traffic.
Complete guide · 38 min
The fourth term: purchase frequency, and why it raises what you can pay for a customer.
Want this run on your business rather than read about it?
We open by naming which term of the growth equation is actually movable at your stage, then build the plan and the cadence around it.