Competitive Analysis for Startups: Research That Changes Decisions

1. Blog
2. Strategy & Positioning

By Saroj Jha · October 1, 2026 · 11 min read

Start with what customers would do if you didn’t exist, which is often a spreadsheet or doing nothing, not only the startups you can name. Then find where you clearly win, and say only that. The five steps below show how, with one example company, and end with the competition slide investors expect.

Too many competitive analyses are feature grids of the three best-known rivals, built once for a pitch deck and never opened again. They miss the biggest competitor, which is usually the status quo, and they rarely change a decision. The point of the research is to decide what to build, say and charge.

Researchers define competitive intelligence as “the process and forward-looking practices used in producing knowledge about the competitive environment to improve organizational performance”^[4]. Competitive analysis is the part of that a startup can do itself, in a few hours a quarter. In my view, it is also the research founders find hardest to turn into an investor pitch, so this post ends with the competition slide.

Running case: Example Co. Fictional

Example Co. is a seed-stage EdTech startup with an AI reading coach for children in kindergarten to grade 3: children read aloud to it, it flags mistakes, and adults get a weekly progress report. It sells first to independent tutoring centers with 2 to 10 tutors. In this post it maps its competition for those centers, and its founders need a competition slide for their next investor pitch. Example Co., its competitors (Competitor A, B and C) and all its numbers are illustrative.

Step 1: List every alternative, not only the rivals you can name

Your competitors are every option a buyer weighs to solve the same problem. There are four kinds:

• Direct rivals: products that do roughly what you do, for the same buyer.
• Adjacent tools: broader products that cover your job as one feature among many.
• In-house workarounds: spreadsheets, scripts, an agency or an extra hire.
• Doing nothing: living with the problem for another quarter.

The last one wins more often than founders expect. Matthew Dixon and Ted McKenna analysed more than 2.5 million recorded sales conversations and found that 40% to 60% of deals are lost to buyers who say they intend to buy but never act^[1]. They argue that many of these deals die of indecision, the buyer’s fear of getting it wrong, and not only of a preference for the old way. Either way, “no decision” is a competitor you have to plan for.

Michael Porter made the same point about strategy in general: managers “define competition too narrowly, as if it occurred only among today’s direct competitors”^[5]. The jobs-to-be-done view adds that customers “hire” and “fire” products for a job, so a spreadsheet the customer already uses is a real competitor^[6].

Large companies say the same in their own filings. ServiceNow’s 2012 IPO prospectus says: “We face competition from in-house solutions, large integrated systems vendors and smaller companies with point solutions including SaaS offerings”^[7]. Slack’s 2019 registration statement named Microsoft as its primary competitor, then added: “We further compete against existing software, programs, and tools, such as email”^[8].

Example Co., step 1 Fictional

The founders ask four center owners and two lost prospects what they use today. The list comes to five options: Competitor A, an AI reading-practice tool sold to schools and larger centers; Competitor B, the scheduling and billing suite many centers already pay for; Competitor C, a school reading assessment; tutor notes in a spreadsheet, with updates sent by email or text; and doing nothing, where parents hear “going well” at pickup. All fictional.

Do this

Ask your last five won and lost prospects, “What would you have done if we did not exist?” Add every answer to the list.

Step 2: Answer six questions about each alternative

For each one, answer six questions, and write down where each answer came from and when:

Complaints matter most. In Forrester’s State of Business Buying, 2024, a survey of more than 16,000 business buyers worldwide, 81% were dissatisfied with the provider they had chosen^[2]. Your opening is usually in the gap between what a rival promises and what its customers experience.

Example Co., step 2 Fictional

The team fills one row per alternative in the Competitor Research Grid. Two complaints stand out: Competitor A’s reports are written for teachers, not parents, and Competitor B’s progress notes are typed by hand. For the spreadsheet, the complaint is tutor time. For doing nothing, owners say families leave when they cannot see progress. All findings are fictional.

Do this

Fill the complaints column first. It is where your opening usually is.

Free template

Profile every alternative in one grid

The free Competitor Research Grid gives each alternative, including workarounds and doing nothing, one row with a source and a date on every entry, plus a change log and a sheet that turns into the investor slide.

Step 3: Gather the evidence from free sources, legally

Most of what you need is public or already in your own conversations:

• Their own pages: pricing, product documentation, changelogs, integration listings.
• Job postings: what a competitor hires for shows where it is investing.
• Review sites and communities: I’d start with the three-star reviews, which in my experience name specific trade-offs.
• Your own deals: ask every won and lost prospect what else they considered and why they chose as they did. The Win-Loss Analyzer helps you code the answers.
• Buying the product: sign up under your own name and company, and read the terms of service first, because some forbid it. Salesforce’s, for example, bars customers from using its service to “build a competitive product or service” or to “copy any ideas, features, functions or graphics of the Service”^[10].

Stay on the right side of the line. US trade-secret law counts theft, bribery, misrepresentation, breaking or inducing someone to break a duty to keep a secret, and espionage as improper means; reverse engineering, independent derivation and other lawful means are not^[11]. In practice: never pose as a customer, a job candidate or a researcher to hide who you work for, and never ask a competitor’s current or former employees, or its customers, for confidential material.

Buyers do much of this research themselves. In a Gartner survey of 646 B2B buyers in August and September 2025, published in March 2026, 67% said they prefer a rep-free experience^[3]. The public signals you study are the same ones they compare.

Example Co., step 3 Fictional

In one afternoon the founder reads the three competitors’ pricing pages and product documentation, sorts their three-star reviews, and replays six recent demo calls, recorded with the centers’ consent, for what else each center considered. The founder does not sign up for any trial whose terms forbid use by competitors. Fictional.

Do this

Put a source and a date next to every entry. An undated entry is a rumour.

Step 4: Sort where you win, lose or it is close, then decide

For each alternative, sort buying situations into three groups: where you clearly win, where you clearly lose, and where it is close. Then build your positioning on the “clearly win” situations and stop spending effort on the “clearly lose” ones.

April Dunford’s method starts in the same place: the competitive alternatives a buyer would use, then the capabilities only you have, then the value those create, then the buyers who care most about that value, and only then the market category, in the updated and expanded edition of Obviously Awesome (2026)^[12]. The description of her book Sales Pitch (2023) promises to show how “do nothing” is “the most fearsome competitor you have”^[13]. My positioning guide turns this into a statement you can use.

Research is only useful if it changes a decision. John Horn worked for nine years on competitive strategy and war-gaming workshops at McKinsey before writing Inside the Competitor’s Mindset (2023)^[14], and he says the point of competitive insight is “to get a better understanding of what might happen in the future”, not to explain the past^[15]. Three decisions it should change:

• Pricing. Buyers anchor on the alternative. Against a rival, price against its package, not its list price alone. Against a workaround or doing nothing, price against the cost of the problem: the hours, errors or lost revenue it causes today. The Pricing Page Teardown template helps you compare packages side by side.
• Messaging. Answer two questions in order: why change at all, which beats the status quo, and why us, which beats the rivals. Most startup messaging answers only the second, which leaves the biggest competitor unaddressed.
• Naming competitors. The FTC encourages comparative advertising that names competitors, as long as it is truthful and not deceptive, with clear wording and, where needed, a disclosure; comparative claims need the same proof as any other claim^[16]. Federal trademark law counts advertising that lets buyers compare products as a fair use of a rival’s mark against dilution claims, but a rival can still sue over an ad that misrepresents its product or suggests it endorses you^[17]. So the question is mostly strategic: naming a larger rival can make you look small or send buyers to it. Compare openly on sales calls and comparison pages; lead your main message with the problem instead.

Example Co., step 4 Fictional

Example Co. clearly wins at centers with 2 to 10 tutors and K–3 readers whose owners have lost families over unclear progress. It clearly loses at large chains that want one suite for everything, and at centers required to use school assessments. It is close at one-tutor centers, where the tutor already talks to every family, so the team tests those in demos before deciding. So the message leads with the problem, “Show every parent their child’s reading progress every week, without extra tutor time”, and the price of $200 a month is set against the cost of losing a family, not against Competitor B. The team also sees that Competitor B could be a partner: centers already pay for scheduling tools like it, so sending reports through those tools is a channel, not only a fight. All fictional.

Do this

Before you file the research, write down the one decision it changes: what you build, say or charge.

Step 5: Turn the analysis into the competition slide investors expect

Investors will ask who else solves this problem, and the slide should answer with the alternatives your buyers actually consider, compared on what those buyers care about. My post on the market and competition slides covers what investors look for. Here is how the research above becomes the slide:

• Rows: every real alternative. Direct competitors, indirect ones, the spreadsheet or manual process, building in-house, and doing nothing. The US Small Business Administration’s guide lists, among the things to assess, “indirect or secondary competitors who may impact your success”^[18].
• Columns: two or three buyer criteria. Take them from your interviews and the complaints column, not from features only you have.
• Cells: plain words, not ticks. Be honest where others are stronger. A slide where you win every row is hard to believe.
• One line on why you will keep winning. Data, distribution, switching costs, a network effect or focus on a segment others ignore.
• The research in the appendix. Put the grid, with sources and dates, in the appendix or data room so an investor can check it.

Never write “no competitors”. Every buyer solves the problem somehow today, even if by doing nothing, and an investor will think of an alternative within seconds.

What Zoom’s filings show

SEC rules ask companies to describe their competitive conditions, where material, when they go public and in every annual report^[19], and Zoom’s filings show the shape of a good competition slide. They show what Zoom said, not how it made its choices internally.

In its 2019 IPO prospectus, Zoom grouped its competitors by type rather than listing logos. It said the market for “communication and collaboration technologies platforms is competitive and rapidly changing”, and that certain features of its platform competed with products offered by “legacy web-based meeting providers, including Webex and Skype for Business”, “bundled productivity solutions providers with basic video functionality, including Google” and “point solutions providers, including LogMeIn”^[20]. It was also candid about where rivals were ahead: many of them “benefit from competitive advantages over us, such as greater name recognition, longer operating histories, more varied products and services, larger marketing budgets”, among others.

Its first annual report after the IPO named the factors it competes on: “We believe we compete favorably based on the following competitive factors”, namely a “video-first platform”, “cloud-native architecture”, “functionality and scalability”, “ease of use and reliability”, the “ability to utilize existing infrastructure, such as legacy conference room hardware” and “low total cost of ownership”^[21].

Six years on, both lists had moved. In its annual report for fiscal 2026, Zoom opens the factors with an “open, AI-first work platform” in place of the “video-first platform”, adds “brand awareness and preference”, and names new kinds of rival: contact-center providers such as Five9 and Genesys, and consumer platforms such as Amazon, Apple and Facebook^[22].

That is the competition slide in miniature: alternatives grouped by kind, a short list of factors buyers weigh, and an honest line on where others are stronger. The fiscal 2026 changes show why yours needs updating as your market moves.

Example Co., step 5 Fictional

The slide has six rows: doing nothing, tutor notes and spreadsheets, Competitors A, B and C, and Example Co. The three columns come from the owner interviews: do parents see progress every week, how much extra tutor time does it take, and does it fit a small center’s budget and tools. Competitor B wins the third column for centers that already pay for it, and the slide says so. The research grid goes in the appendix. All fictional.

Do this

Draft the slide from your grid, then ask someone who was not in the research to find a row where you lose. If they cannot, the slide is not honest yet.

How often should I update it, and who owns it?

Give it one owner, the founder at first, and two rhythms:

• A light quarterly review: pricing, positioning and new entrants.
• Event-triggered updates: when a rival changes pricing, launches something relevant, or appears in several lost deals in a row.

Keep a dated change log, so you can see how a competitor moves over time rather than only where it stands today. The Competitor Research Grid template includes one. Update the competition slide from it before each fundraising round.

What changes by stage?

• Pre-seed — Focus on the workarounds and doing nothing. Your first customers are mostly choosing between you and the status quo, and your competition slide should show what they used before you.
• Seed — Add the two or three rivals that appear in your deals. Build the grid, start logging why you win and lose, and put that win-loss evidence on the slide.
• Series A–B — Formalise it: one owner, the quarterly review, battlecards for sales, a comparison page for each rival buyers ask about, and win rates by competitor on the slide.

What mistakes should I avoid?

• Copying features. A rival’s roadmap reflects its customers, not yours.
• Ignoring doing nothing, a common way to lose.
• Researching only the loudest rival instead of the ones in your deals.
• Attacking by name in your main message.
• Building it once and never updating it.
• Writing “no competitors” on the slide. Every buyer uses something today.
• A slide you win on every row. It reads as marketing, not analysis.