Competitive positioning: frameworks & methodology

A data infrastructure company can plot itself on three maps in the same week and get three different answers. Against on-premise warehouses it is cheap and elastic. Against open-source query engines it is expensive and closed. Against the analytics tools its buyers already own it is barely a competitor at all. Same product, three positions. The frame you pick decides which comparison the buyer runs, which budget line pays you, and what price you can hold. The models below exist so that choice rests on evidence: perceptual maps to see where buyers already place you, value curves to see what you would have to give up, and category tests to check whether the frame you want can carry a purchase order.

Choosing the frame before you draw anything

Take the frame of reference as the foundations lesson sets it out, and treat choosing one as a decision with a price attached rather than the output of a workshop. Two questions do most of the work. The substitution question, asked of buyers who signed in the last six months: if we had not existed, what would you have done instead? The answers cluster into two or three alternatives, and that cluster is your competitive set, whatever your deck claims. Then the budget question: which line item paid for us? A product funded from an existing warehouse budget gets compared to warehouse prices. A product with no line item needs a champion to invent one, which adds weeks to every deal and puts an unfamiliar name in front of a finance committee.

The two answers have to agree with each other and with the difference you claim. If the buyer is a group CFO, the frame is business intelligencebusiness intelligenceTechnologies and processes that turn raw data into actionable insights via reporting, dashboards and analysis, so teams can decide based on facts rather than intuition.View full definition →, and the claim is ease of use, the pieces pull apart: that buyer weights audit trail and ERPERPA single integrated software backbone that runs core operations: finance, procurement, supply chain, HR and manufacturing on shared data.View full definition → integration, and the ease-of-use line gets nodded at and forgotten. Mismatches of this kind rarely surface as an objection you can hear. The deal just goes quiet in week five.

Framework 1: perceptual mapping

Plot yourself and the real alternatives on two axes buyers use, derived from interviews rather than from a whiteboard. Ten to fifteen win/loss conversations is usually enough; by the twelfth, criteria stop being new. Take the two criteria mentioned most often at the decision point, then place every player using evidence a buyer could verify.

Three ways the mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → lies to you. First, correlated axes: if you plot price against quality, every vendor lands on a diagonal, the map shows a ranking and no white space. Pick axes that can genuinely vary independently. Second, axes chosen because they flatter you, which produces a chart nobody outside marketing recognises. Third, and most expensive, the empty quadrant that is empty for a reason. Before you call it white space, name three buyers who asked for that combination and state what they said they would pay. Plenty of quadrants are vacant because the economics do not work, not because nobody thought of it.

Framework 2: value-curve comparison

The strategy canvas from W. Chan Kim and Renée Mauborgne at INSEAD gives you more resolution than two axes. List the six to ten factors the category competes on, score each player one to five from external evidence (pricing pages, review sites, RFP responses, not memory), and draw the lines.

Read the shape, not the scores. If your curve tracks the leader's within a point on every factor, you are selling the same thing and the negotiation will end on price. A real position looks like a line that drops hard somewhere and rises hard somewhere else. The working discipline: every raise must be funded by a cut. Eliminate a factor, reduce another, raise a third, create something the category does not offer. If nothing on your curve sits below the field, you have drawn a wish list.

Framework 3: category-definition tests

Before you name a category, run four tests. Is there a budget line, and if not, whose budget gets raided and who signs? Do buyers already use words for this without promptingpromptingPrompt engineering is the practice of designing and refining text inputs to guide large language models toward accurate, relevant, and reliable outputs.View full definition →, in inboundinboundA strategy that attracts prospects organically via valuable content (blog, SEO, social) rather than interrupting them.View full definition → enquiries and in job postings? Will procurement drop you into an existing comparison grid anyway (if the RFP template exists, you are in that category whatever you call yourself)? And can a customer describe you to a colleague in one sentence without using your brand name?

Snowflake is the useful counter-example, because the technology was new and the frame was old. Separating storage from compute across AWS, Azure and Google Cloud was a genuine architectural break, but the company came to market as a data warehousedata warehouseA central repository that consolidates data from many source systems into a structured, query-optimized store designed for analytics, reporting, and business intelligence.View full definition → built for the cloud. That frame came with a budget line, a familiar comparison against Teradata and Oracle, and no need to teach anyone why a warehouse exists. Only later, around 2020, with scale behind it, did Snowflake put a name of its own on the table with the Data Cloud. Its September 2020 listing was the largest software IPOIPOThe first sale of a private company's shares to public investors on a stock exchange, converting private ownership into publicly traded stock.View full definition → to that point, and product revenue passed $2.5 billion a year by the fiscal year ending January 2024. Borrow the frame first, name the category once you can afford the teaching.

Framework 4: costing the frame before you commit

Give each candidate frame a rough profit and loss. An inherited frame hands you a price ceiling set by the incumbent, shorter cycles, and a features checklist you must satisfy to stay on the grid even where those features do not matter to your buyer. A new frame has no ceiling and no grid, but you pay for the education, and third parties (analysts, review sites, buyers' own vocabulary) take a couple of years to adopt a word, assuming they ever do. Quantify it crudely: estimate the extra weeks per deal, multiply by expected deal count, and compare that to the pricing headroom the new frame buys. Second-order effect worth naming: an inherited frame also inherits the incumbent's objection list, so your reps will spend calls answering for capabilities you decided on purpose not to build.

Real-world case: ikea

Score the furniture retail category on price, design, sales assistance, delivery, assembly, in-stock availability, store convenience and range, then draw IKEA against a traditional retailer. Assistance drops close to zero in the self-serve warehouse. Delivery and assembly move to the customer, a choice that dates to the flat-pack experiments of the mid-1950s after Gillis Lundgren took the legs off a table to fit it into a car. Stores sit out of town on cheap land, so convenience falls too. Those four cuts pay for the raises: design at a price point the category did not offer, stock you walk out with the same day, and range under one roof. IKEA then created factors the category had no line for at all, including the fixed showroom route, the restaurant and childcare. The result is retail sales above €40 billion a year and hundreds of millions of store visits. What transfers is not the furniture. It is that the concessions are visible to the customer and priced in, and a curve without cuts is not a position.

Real-world case: segment

Segment shipped one APIAPIApplication Programming Interface: a standardised interface that lets applications communicate and exchange data without knowing each other's internal workings.View full definition →: collect customer event data once, route it to hundreds of destinations. The obvious frame in 2013 was analytics, and the founders refused it. The consequence ran through the whole business. Every analytics vendor became a destination, which turned potential rivals into a partner list and a distribution channel. The cost was that no buyer had a budget line for customer data infrastructure, so the company sold to engineers rather than to marketing leadership, open-sourced analytics.js in December 2012, and let adoption happen below the level where procurement needed a category name. When analysts later formalised the customer data platform label, Segment was already inside the grid as an incumbent rather than fighting to be added to it. Twilio bought the company in a deal announced in October 2020 worth roughly $3.2 billion in stock. Set that against Snowflake and the two orderings are clear: one borrowed an existing budget line and named its category afterwards, the other refused the existing frame and waited for the market to name a category around it. Both work. Choosing neither, and describing yourself in a category you neither own nor fit, does not.

Dunford sells positioning consulting, so treat the method as a vendor's method and test it against your own win/loss data.

CMO action items

• Rebuild the perceptual map each quarter from fresh interviews, not from last quarter's file. If your point overlaps a competitor with more awareness or a lower price, the map has told you something and the answer is to change the frame or change the product.
• Ask five lost deals per quarter the substitution question and record the exact category word the buyer used for you. Compare it to the word in your homepage headline. Drift between the two is the earliest signal that the field has quietly re-framed you.
• Write the positioning brief on one page: buyer, frame, the factors you raise, the factors you cut, and the price ceiling that frame implies. Cuts and ceiling are the parts that get deleted in review, and they are the parts that make the document worth having.

Common mistakes that kill positioning

• Positioning by committee. When the statement has to satisfy sales, product, finance and the founder, it turns into a feature list in aspirational clothing. Somebody has to hold the authority to refuse additions.
• Differentiating on a factor buyers already assume. Claiming security in a market where every shortlisted vendor is SOC 2 audited is describing table stakes. The value curve catches this: if all lines sit at four or five on a factor, it cannot carry your difference.
• Changing the frame every two or three quarters. Market perception moves on repetition, and 12 to 18 months of consistent naming is the usual minimum before third parties use your word back at you. Repositioning inside that window resets the clock and leaves sales with three vocabularies in circulation at once.