Strategy & Growth · Guide

How to build a business strategy, decision by decision

Eight guides rank for this question and all eight answer with steps. Not one of those steps is a decision, and not one of them dates a framework it tells you to use.

Reviewed August 2026 · The Insight Journal Editorial Team

In short

How to build a business strategy, in one sentence: convert one chosen growth lever into a written document that names a target with a number and a date, the budget that funds it, the evidence that would stop it, and the cadence on which it is reviewed. Five decisions, five artifacts, usually one page. The lever itself comes from the four levers and the 1957 paper they come from.

Building a business strategy: a printed strategy document, sand-coloured folders and a fountain pen on an oak desk in raking morning light.

The artifact

What a finished business strategy has to state

In short

A finished strategy states six things: the lever, the refusal, the target, the budget, the stop rule and the review cadence. Each is one line, and each is a decision somebody has to make out loud. A document missing any of the six is a description of the market rather than a strategy.

01

The lever

One named route to more revenue, chosen over the others on purpose.

02

The refusal

What the company is therefore not doing this year, written where staff can read it.

03

The target

One number with one date, measured against a baseline that exists today.

04

The budget

What the attempt costs to run, and what it costs if it fails.

05

The stop rule

The evidence that ends it, agreed before any money moves.

06

The cadence

The review date, and the name of the person who chairs it.

The distinction that matters here is small and load-bearing. A strategy is the choice of lever. A plan is the budget, target, date and stop rule attached to that choice.

Most published advice on how to build a business strategy stops at the first thing. This page builds the second, and hands the taxonomy question sideways rather than restating it.

Two of the six statements are the ones nobody writes. Almost every strategy document names a target. Almost none names a refusal or a stop rule, which is why so many of them quietly outlive their own evidence.

Provenance

What the guides on how to build a business strategy leave out

In short

All eight pages ranking on this query in the United States on 18 August 2026 answer with a numbered list, using five different step counts between them. Every step is additive. None asks what the company will stop doing, attaches a cost, names a threshold for abandonment, or dates a single framework it recommends.

8 of 8

Ranking pages answer with a numbered list of steps, using five different step counts

Live US results pull, 18 August 2026

0 of 8

Date, author or cite a single framework they instruct the reader to use

Live US results pull, 18 August 2026

7 of 8

Are published by an organisation selling into the category, none of which discloses it

Publisher about-pages, checked individually

2,174

Median measured word count of the four pages we crawled in full

Live crawl, 18 August 2026

The live United States ranking set for this query, what each publisher sells, and what its guide contains
Publisher	What it sells	What the page is
IMD	Executive education, from a business school in Lausanne	5 build steps, 2,141 words
Harvard Business School Online	Online certificate courses	6 steps, dated 25 October 2022
The Strategy Institute	Paid strategy certifications	5 steps, 2,206 words, every body heading at H4
Vistage	CEO peer-advisory memberships	10 steps, 1,857 words, closes on a membership pitch
US Small Business Administration	Nothing. The only such result in the set	The lean one-page plan and the break-even formula
skillsandbusinesshub.org	UK skills and apprenticeship services	3,565 words, five steps mirroring IMD, UK spelling
Simon-Kucher	Pricing and strategy consulting	Answers business development, a different query
Business Development Bank of Canada	Financing and consulting, in Canada	Ranks on a United States query

Three things worth knowing

Two publishers in a United States results set are not United States publishers. One is a UK skills hub whose theme generator string still reads RENAME ME v.1.0.0, and one is a Canadian bank.

The sixth result's five steps mirror the first result's five steps almost line for line. Two pages bury every content heading at H4 with no H2 parent above it.

Our own disclosure

The Insight Journal earns money from sponsored articles and advertising. It sells no courses, certifications, memberships, consulting or financing, so nothing below routes to a product we own.

Every framework named on this page carries its publication and year, and how we research, cite and disclose sets out how that is enforced.

Decision one

Pick one lever, and write down what you are therefore not doing

In short

Choosing a growth lever is a decision to spend the year on one route to revenue instead of the others. The artifact is two lines: a one-sentence commitment, and a refusal list naming the things the company will not attempt while that commitment stands. The second line is the one that makes the first real.

Strategy is subtraction before it is anything else. A firm with one lever and four refusals has made a choice. A firm with four levers has made a wish list and will fund none of them properly.

Write the refusal where staff can read it. An unwritten refusal gets quietly reversed in the third month by whoever is closest to the customer asking for the thing you decided not to build.

The route itself is a taxonomy question rather than a procedural one, and which lever the matrix says you are choosing between works through the four options and the trade-offs between them.

The artifact

Commitment

"This year we grow by selling more of the existing product to the buyers we already have."

Therefore not

No second product line.
No new geography.
No acquisition conversations.
No channel partner we would have to train.

Four refusals is a working number. One is not a choice, and ten is a mood.

Decision two

Turn it into one target with a number and a date

In short

A target is one metric, one figure and one date, measured against a baseline that already exists. One target, not a scorecard: a plan with nine measures has no priority, and the first month of conflict resolves it in whichever direction is most comfortable. If the baseline cannot be produced today, the target cannot be judged later.

SMART, as published

George T. Doran published the SMART criteria in Management Review 70(11) in November 1981. In the original, A means assignable and R means realistic.

The version every ranking guide repeats gives achievable and relevant. That drift removes the two letters with teeth: somebody's name against the target, and an honest view of whether the firm can hit it.

S specific
M measurable
A assignable
R realistic
T time-related

Write the baseline first and date it. "Repeat order rate was 21% in the quarter to 30 June" is a baseline. "Repeat orders are low" is a feeling.

Then attach the assignable letter. One person's name sits next to the target, and that person is the one who reports the number at each review, whether it moved or not.

Pick the date before picking the figure. A twelve-month horizon invites a round number nobody has to defend, while a two-quarter horizon forces an estimate somebody can actually reason about.

Decision three

Budget it honestly, including the cost of it failing

In short

Budgeting a strategy means costing six lines, not one. Direct spend is the line everybody writes. Tooling, management attention, displaced work, working capital and the cost of unwinding are the five that decide whether the firm can carry the attempt, and whether it can walk away from it.

The six budget lines a strategy needs, what each covers, and why it matters
Line	What it covers	Why it is on the list
Direct spend	Cash that leaves the account to run the attempt	Usually the only line the guides mention
Tooling and subscriptions	Software, listings, agency retainers, anything recurring	Recurs after the attempt stops unless somebody cancels it
Management attention	Hours of the owner or a manager, priced at what those hours otherwise produce	The scarcest line in a firm with no planning function
The displaced work	What stops getting done because the strategy is being run	Never zero, and rarely written down
Working capital	Cash tied up between the spend and the first receipt	Sets how long the firm can hold the position
The cost of unwinding	Exit fees, notice periods, written-off stock, an unhappy customer	The line that decides whether the decision is reversible

Structural framework from The Insight Journal. No dollar figures are printed because they are firm-specific and no primary source publishes a benchmark for them.

Two arithmetic questions finish this decision. How many units, hours or contracts does the attempt have to produce before it covers itself, and how long can the firm hold the position while that happens.

The first is a break-even calculation: fixed costs divided by the sales price per unit minus the variable cost per unit. The US Small Business Administration publishes both that formula and a lean one-page plan format on Plan your business.

The second is a cash question rather than a strategy question, and putting the spend into a weekly cash forecast is where the answer comes from.

Decision four

Decide now what evidence would make you stop

In short

A stop rule is a threshold, a source of evidence and a date, written before the money moves. It exists because the judgement gets worse as the sunk cost grows. Agreeing in month zero that a given reading ends the attempt is cheap, and agreeing it in month seven is nearly impossible.

The four stop-rule categories, with the threshold, the evidence and the review date for each
Category	Threshold	Evidence	Checked
Cash	Cumulative spend passes the figure agreed in decision three	Management accounts	Month 3
Demand	The baseline metric has not moved at all, in either direction	The same report the baseline came from	Month 3
Delivery	The existing business misses commitments it was meeting before	Whatever already measures delivery	Reviewed monthly
Attention	The named owner has missed two consecutive review meetings	The calendar	Every review

Categories and structure from The Insight Journal. The thresholds shown are shapes to fill in with your own figures, not benchmarks: no primary source publishes a defensible abandonment rate for this.

A stop rule is not a target miss

Missing the target by a margin at the deadline is information about the estimate. Hitting a stop-rule threshold is information about the lever. They call for different conversations.

Tie it to the payback period

The clock the rule runs against is the payback period from decision three. A threshold with no date attached gets read as a suggestion by everyone who wants the attempt to continue.

Watch the delivery line

The existing business is the first thing a growth push damages, and the mechanism by which fast growth cracks a business sets out how that damage builds before revenue shows it.

Decision five

Set the review cadence and name who owns it

In short

The cadence is a recurring date in the calendar with one name against it, not a meeting that happens when results turn bad. Quarterly fits most firms carrying a single strategy. The standing agenda is four items long and the same four items every time, which is what makes the meeting short enough to survive a busy quarter.

01 The target against the baseline, as one number.
02 Cumulative spend against the budget agreed in decision three.
03 Each stop-rule threshold, read out and answered yes or no.
04 Whether the refusal list still holds, or something crept back in.

One name, not a committee. A strategy owned by the leadership team is owned by nobody, and the first quarter in which the number is bad is the quarter the meeting gets moved.

Give the owner the authority that goes with it. The person reporting the number should be able to call the stop rule without asking permission, which is the only thing that makes decision four binding.

It is a demanding role in a small firm, and whoever has to chair that meeting every quarter covers what the job asks of them.

One more check belongs on the agenda early: whether the operation can absorb the work the plan creates. A target the delivery side cannot serve is not a strategy, it is a complaint scheduled for later.

The highest-value table on this page

Where the frameworks these guides name actually come from

In short

Five of the frameworks named in this category have an author, a publication and a year. Not one of the eight ranking pages gives any of them. Two are also misquoted almost universally: SWOT began as SOFT, and Doran's original SMART letters are assignable and realistic.

Each framework with its author, original publication, year, and the correction it needs
Framework	Author	Published in	Year	What the guides get wrong
SWOT analysis	Robert Franklin Stewart	Long Range Planning 56(3), article 102304	2023 study of 1960s origins	Began as SOFT. The widely repeated Albert Humphrey story is not what the archival research found.
SMART criteria	George T. Doran	Management Review 70(11)	November 1981	A is assignable and R is realistic. Achievable and relevant are later drift.
Balanced scorecard	Robert Kaplan and David Norton	Harvard Business Review	January and February 1992	Four measurement perspectives, published as a measurement system rather than a planning method.
Five forces	Michael E. Porter	Harvard Business Review	March and April 1979	Extended in Competitive Strategy, Free Press, 1980, which also carries the generic strategies.
Objectives and key results	Andrew S. Grove	High Output Management	1983	Documented as Intel's version of management by objectives. The acronym was popularised later.

The SWOT correction

The archival study The origins of SWOT analysis, published in Long Range Planning in 2023, credits Robert Franklin Stewart and finds the technique started life as SOFT. The Albert Humphrey story repeated across strategy blogs is not what the record supports.

The other three citations are straightforward. Kaplan and Norton published The Balanced Scorecard: Measures That Drive Performance in 1992, and Porter published How Competitive Forces Shape Strategy in 1979, both in Harvard Business Review.

Three lists with no owner

The 5 C's of strategic management

No traceable original publication. The letters differ between the sites that list them, which is what an unowned framework looks like.

The five pillars of strategy

Same pattern. Widely listed, never cited, and the pillars themselves change from page to page.

The 1% rule in business

Used for at least three unrelated ideas: compounding improvement, a marketing spend ratio, and the share of an online audience that posts. None has an owner.

If there is no planning function

The one-page version

In short

Five sentences on one page is a complete strategy for a firm without a planning function. Write the lever, the refusal, the target with its date, the budget with its worst case, and the stop rule with its threshold. The SBA publishes a lean one-page plan format that takes about an hour to fill in.

01 We will grow this year by [lever], and not by [refusals].
02 Success is [metric] moving from [baseline, dated] to [figure] by [date].
03 [Name] owns that number and reports it every quarter.
04 It costs [direct spend] plus [hours] a week, and [figure] if it fails.
05 We stop if [threshold] on [evidence] by [date].

Questions

Building a business strategy: common questions

How do I create a business strategy?

Make five decisions and write each one down. Choose one growth lever and record what you are therefore not doing, set one target as a number with a date, budget what the attempt costs to run and to fail, agree the evidence that would stop it, and put a review date in the calendar with one named owner. The document that results is the strategy. Anything shorter is an intention.

What is the difference between a business strategy and a business plan?

The strategy is the choice. The plan is the arithmetic attached to it: the budget, the target, the date and the stop rule. A business plan in the SBA sense is also the document you show a lender, which is why it carries market and financial sections a strategy does not need.

How long should a business strategy document be?

One page is enough for most firms, and the SBA publishes a lean one-page format for exactly this. Length is not the quality signal. A page carrying a lever, a refusal, a number, a date, a budget and a stop rule beats thirty pages of market description with no decision in them.

What are the 5 P's of business strategy?

Plan, ploy, pattern, position and perspective, from Henry Mintzberg in California Management Review 30(1) in 1987. They are five meanings of the word strategy, not five components of a plan, which is how most pages repeat them. It is the only one of the numbered lists Google suggests for this topic that has a traceable published source.

What are the 5 elements of business strategy?

There is no sourced list of five. Different publishers name different elements and none cites an original. If you want a defensible set, use the six statements at the top of this page: lever, refusal, target, budget, stop rule and cadence, each of which produces a written artifact.

What are the 5 C's of strategic management?

No original publication is traceable for this one. The letters vary between the sites that list them, which is the signature of a framework nobody published and everybody repeats. Treat it as a memory aid rather than a method.

What are the five pillars of strategy?

Another unowned list. It appears widely, is never cited, and the pillars themselves change from page to page. Where a framework has no author and no date, it also has nothing to check.

What is the 1% rule in business?

It refers to at least three unrelated ideas: improving something by one percent repeatedly, spending one percent of revenue on a given line, and the share of an online community that creates content. None is attributable to a published source, so name which version you mean before using it in a plan.

Who invented SWOT analysis?

The 2023 archival study in Long Range Planning credits Robert Franklin Stewart and finds the technique began as SOFT rather than SWOT. The commonly repeated story that Albert Humphrey invented it at SRI is not what that research concluded. It is the best-evidenced answer currently available.

How often should a business strategy be reviewed?

On a fixed date rather than in a crisis, and quarterly suits most firms carrying one strategy at a time. What matters more than the interval is that one named person chairs it and the same four agenda items are answered every time. A review that only happens when results are bad is not a cadence.