Tarot competitor analysis: practical guide, examples and limitations

Tarot can support competitor analysis in one limited but useful way: it can help a team form hypotheses about positioning and notice questions it has been avoiding. It cannot reveal a rival’s private strategy, intentions or future moves. Start with public facts—product flows, pricing, reviews, help pages, promises, release notes and visible constraints. Then use the cards to explore more than one explanation: what truly differentiates you, what a competitor does better, where the category is overcrowded and which customer need may be underserved. Every interpretation must return to open data, interviews or a small positioning experiment. Without that check, the reading is a compelling story, not strategy.

I once covered a whiteboard with competing features. The comparison was accurate, yet the longer we stared at it, the more we wanted to copy the other product’s buttons and claims. The useful shift came from a different question: “What job are customers still struggling to complete, even with both products available?” The cards did not provide secret information. They changed the angle of inquiry, which was exactly what the team needed.

Facts come before cards

Competitor analysis begins by separating what can be observed from what is merely inferred. A landing page shows a public promise; it does not prove the company’s internal strategy. Reviews describe particular experiences; they do not represent every customer. Job adverts may suggest an area of investment, but they do not guarantee a future feature.

Build a compact evidence matrix before the reading:

Create the same list for your own product. Otherwise the reading can become a discussion of a rival’s strength and your team’s anxiety. “How do we beat them?” is usually a weak question. “Which difference can we honestly prove, and for whom does it matter?” is much more workable.

Save the evidence and frame the question in Reflecta

Ten positioning and blind-spot questions

Use ten positions defined before you draw. Do not add cards simply because one answer feels uncomfortable.

1. Which real difference do we already possess? Look for a feature, process or relationship with customers, not a slogan.
2. What does the competitor genuinely do better? The position only works if you can acknowledge an advantage without dismissing it.
3. Why do customers choose that competitor now? Treat the card as a hypothesis to compare with reviews and behaviour.
4. Which part of the market is overheated? Where does nearly everyone repeat the same promise?
5. Which customer need is underestimated? Consider clarity, support, control, privacy, confidence or continuity after the first result.
6. What price do customers pay for existing solutions? It may be time, complexity, anxiety, inflexibility or extra work.
7. Where are we likely to imitate instead of understand? Which visible element attracts us mainly because the market leader uses it?
8. Which existing strength are we failing to communicate or use? The difference may already exist but remain hidden in the journey.
9. What small and reversible positioning experiment can we run? Test an explanation before changing the whole brand.
10. Which evidence would make us revise the interpretation? A useful hypothesis must be capable of being disproved.

The final question protects the exercise from confirmation bias. If no possible evidence could weaken your favourite reading, you are defending a belief rather than conducting analysis.

Read the spread without inventing a rival’s motives

Read each card through its assigned position first. The Magician in “what the competitor does better” might point to a clear first step, a fast demonstration of value or the ability to combine several functions into one understandable flow. It does not prove that the company has a hidden plan or a particular intention.

Write at least two readings. The Hermit in “our difference” could represent depth, privacy and room to think without noise. It could also show excessive isolation, a slow start or a product that asks for patience before offering value. Both readings remain plausible until evidence makes one stronger.

Use this protocol:

1. Describe the visible action, tension or relationship in the image.
2. Connect it only to the named position.
3. Write two plausible hypotheses.
4. List evidence for and against each one.
5. Name one public source or experiment that could test it.
6. Do not infer the private motives of people inside another company.

Tarot is useful when it widens the set of questions. It becomes poor strategy when a vivid symbol replaces customer and market research.

Worked example: a service in a crowded category

This is a composite case. A small team is building a digital service that helps people reflect on difficult situations. A larger competitor offers a very fast start, bold promises and many automated suggestions. The team fears that a slower, more deliberate experience will disappear in the noise.

Known facts:

• the competitor produces a first result in under a minute;
• public reviews often praise the simple and impressive opening;
• some reviews say the next step is unclear after the first answer;
• the smaller service asks for more context but lets a user save the situation, return later and compare the conclusion with a real event;
• the team does not yet know whether target users see this depth as value or friction.

The spread shows the Hermit for the team’s difference, the Magician for the competitor’s advantage, Six of Wands for the overheated zone, Queen of Pentacles for the underserved need, Seven of Swords for imitation risk and Page of Pentacles for the experiment.

Reading one. The team’s difference is quiet depth and a private pace. The rival is better at showing value immediately. The crowded area is the promise of instant insight. The neglected need is stable support after the initial moment. The imitation risk is adopting the rival’s speed and voice while losing the product’s sequence. The experiment is a concise page built around “save a conclusion and test it through an action.”

Reading two. The Hermit may warn that the product leaves users alone for too long. The Magician may represent not only promotion but a better-designed first-use flow. Queen of Pentacles then points to a need for a tangible outcome, and the Page suggests a smaller first step. The experiment becomes a shorter onboarding path that retains the calm tone rather than copying the competitor’s language.

The readings do not fully cancel each other. The team may preserve depth and still make the first result clearer. The spread has produced two testable product and positioning changes, not a mystical verdict.

What a competitor does better is not an instruction to copy

Admitting a rival’s advantage is uncomfortable. Teams often defend themselves with one of two reactions: “they only win because they have more money,” or “we need the same feature immediately.” Neither reaction explains the mechanism.

Break the advantage into four layers:

• Outcome: what the customer obtains.
• Mechanism: which step makes the outcome easier.
• Proof: why the promise feels credible.
• Cost: what the customer gives up for the convenience.

Suppose the rival guides users through the first screen more effectively. Copying the visual design is neither necessary nor appropriate. Instead, identify what reduces uncertainty: one question rather than five, a filled example, visible progress, or a clear privacy explanation. You can learn the principle without reproducing the interface, text, code or protected creative work.

The card in this position should lead to the question, “Which principle creates the advantage?” It is not permission to appropriate intellectual property.

Find the overheated zone and the underestimated need

An overheated zone is an area where most players repeat almost the same promise. The promise may still matter; it simply no longer differentiates. In AI-related products, common crowded claims include “instant,” “personal,” “smart” and “effortless.” Confirm crowding by reviewing public pages and adverts, not by relying on the team’s impression.

An underestimated need often sits next to the cost of the popular promise. When everyone sells speed, some customers may need control. When everyone sells limitless choice, they may need a useful boundary. When communication focuses on inspiration, a quieter group may want continuity and a way to verify what changed.

Ask:

• What does the user need to do after receiving the first result?
• At which point are they left to solve the problem alone?
• Which complaints repeat but rarely appear in positioning?
• Which need can our product genuinely serve rather than merely name?
• What product behaviour proves the promise?

An underserved need is not a guaranteed profitable niche. It is a hypothesis about tension in the current experience and must be tested.

A positioning experiment and an open-data check

A good experiment is reversible, narrow and designed to measure understanding rather than clicks alone. Do not rebuild the brand after one reading. Prepare two honest explanations of the same product.

Show both versions to a small group of target users. Do not ask only which one they like. Ask them to explain, in their own words:

1. Which situation the product is for.
2. How it differs from alternatives they know.
3. What result they expect after the first use.
4. What they would do next.
5. Which claim makes them cautious.

Combine the responses with public information: competitor pages, pricing, help centres, demonstrations, reviews, release history and job listings. Use no confidential material and never label an inference as a fact.

Run the spread and save competing hypotheses in Reflecta

Common mistakes and limits

• Starting with cards instead of evidence. Anxiety can then masquerade as analysis.
• Asking what the competitor is secretly planning. A reading cannot verify private intentions.
• Treating every strong card as your advantage. A position may reveal a weakness or cost.
• Copying a visible feature without understanding its mechanism. What is visible may not be the reason it works.
• Ignoring your own delivery limits. Positioning must be supported by product and operations.
• Testing many changes at once. You will not know what altered the response.
• Treating public reviews as complete market truth. Open data is useful and incomplete.
• Redrawing until a pleasant interpretation appears. That strengthens confirmation bias.

Tarot does not replace customer research, market analysis, legal checks around branding or a financial model. It can help name a blind spot, but responsibility stays with the team.

Action, criterion and review date

Reduce the work to one record:

• Hypothesis: the difference or underestimated need being tested.
• Evidence: what currently supports it.
• Alternative: another credible explanation.
• Action: one reversible experiment.
• Criterion: an observable result that strengthens or weakens the hypothesis.
• Date: when the team will review the result.

For the example above, the action is to show two positioning versions to fifteen target users. The criterion is that at least ten can explain the difference without prompting and at least eight can name the next step in the product. The review date is two weeks later. A weak result means revising the wording or the hypothesis—not drawing another card for reassurance.

This keeps competitor analysis accountable. The cards help reveal a question. Open evidence and an experiment show whether the answer survives contact with reality.

Open Reflecta and record the next testable step

— Leon Maren, official runologist at Reflecta