Skip to main content

The Analyse stage

The Analyse stage decides whether a product is worth entering the market with. It carries eight research modules and produces the opportunity score, the profit model and the entry recommendation. Later stages read its conclusions.

Move between modules with the tabs across the top of the stage.

TabModuleAnswers
01 OverviewOverviewHow good the opportunity is, and how much of the assessment is backed by data
02 CompetitorsCompetitor poolWho the product competes with, and whether the pool is deep enough
03 Market sizeMarket sizeHow large the category is, how concentrated, and which way it is moving
04 KeywordsKeywordsWhich search terms carry the demand
05 VOCCustomer voiceWhat customers complain about. Produced in the VOC stage
06 Profit modelProfit modelWhat the unit earns at each price and cost
07 RiskRisk registerWhat has not been checked yet
08 ConclusionEntry recommendationGo, Conditional Go or No-Go, and what holds it down

How Slate reports a missing figure

These rules apply across every module.

DisplayMeaning
A hatched bar with Not measuredNo data for this figure. It is not a value of zero
0A measured zero
Provisional in amberThe number is correct, and under half the framework carries data
UNVERIFIEDA fee or tariff constant dated 2024 that has not been re-verified
Operator inputA value you entered, used exactly as typed
Not modelledA charge left out of the calculation. It does not mean the charge is zero

A missing sub indicator is never scored zero. It is dropped from the calculation, and the remaining weights are rescaled.

01 Overview

The overview carries the opportunity score out of 100, its four dimensions, the framework coverage and the evidence confidence.

DimensionWeightReads high when
Market demand30%Sales scale, search demand, trend and breadth are strong
Competitive opportunity25%The market is easy to enter. Stronger competition scores lower, not higher
Product fit20%The specification already meets what the market asks for
Profit potential25%Margin, break-even ACOS, downside case and price buffer hold up

Evidence confidence is reported separately and is not part of the score. It states how much of what the conclusion rests on is backed by reliable data or human verification. It does not state how likely the answer is to be correct.

Next best actions lists what is currently unmet. Each entry carries a Go link to the module that fills it.

02 Competitors

Select Generate to collate similar products from the web and the marketplaces. No upload is required.

Slate reports how many rows are real product pages. Search pages, category landing pages and unpriced pages are tagged not a product page and are excluded from the coverage counts. To add a competitor precisely, paste its link into the stage conversation.

Tiers

Assign each competitor a tier. Slate suggests a tier from price alone and marks the suggestion for you to confirm. Price cannot separate a direct competitor from a substitute.

TierTarget depth
Direct10 to 15
Benchmark5
Substitute3 to 5

The coverage bar states the gap in full, for example direct 6 against a target of 10 to 15, four short.

Output

ActionProduces
Titles + links (Markdown)A list of the selected rows
Titles + links (text)The same list as plain text
Excel dataset (.xlsx)A spreadsheet of the selected rows
Enhanced datasetA spreadsheet with extracted specifications. Runs as a background job
Generate ReportA gap report written to the project files

03 Market size

A project that has not been researched shows what a run would produce and offers Run market research. Select it, then select Refresh once the job finishes. The page never claims the work is finished before it is.

SectionContents
Browse node decisionThe category node the research settled on, and the candidate nodes considered
Market size, concentration and trendThe measured figures, each with a source and a reliability tag
Sample validationThe checks behind the confidence grade
Key findingsThe conclusions drawn from the above

The node decision is worded two ways, and they are not equally strong. Several high selling competitors clustering in one node means the node is the real category. Competitors dispersed across nodes means Slate took the densest node, which is a weaker call and worth checking by hand.

ReliabilityTreat as
HighUsable as it stands. Still worth confirming against the sample checks
MediumAn estimate. Confirm the definition and the order of magnitude
LowIndicative only. Verify by hand before any conclusion rests on it

The full report and its raw data tables are written to the project files. The page carries only the key findings and a link reading Open the full report. Figures are never compared across marketplaces.

A run that finishes without obtaining a single figure says so and names the likely cause. Blank is shown in place of invented numbers.

04 Keywords

Filled by the same market research run as 03 Market size. It lists the primary search terms with their monthly searches, and the demand and trend figures that rest on them.

These terms decide which competitors count and which category is the real one, and they are where the advertising and copy work starts.

05 VOC

Review themes and return risk are produced in the VOC stage. Open the VOC stage to upload the review and returns exports, or to read the product problem report already produced there.

06 Profit model

Every figure here is calculated deterministically. No model is involved.

Price bands

Price bands appear automatically once similar products are collated. Each band carries a review weighted density and its representative listings. Select Use this band to apply it.

Cost scenarios

  1. Enter the minimum and maximum unit cost.
  2. Adjust the price and the conversion rate.

Slate reports the unit cost, margin, break-even ACOS and viable CPC under a Downside, Base and Upside case. Where cost exceeds price the cards read Not viable instead of printing a negative bid.

Fee inputs

Leave the fee fields blank to use the 2024 constants, which are marked UNVERIFIED. A value you enter is used exactly as typed and is labelled Operator input in the scenarios and in the report.

FieldCovers
FBA fee (per unit)Fulfilment cost per unit
Referral rate (0-1)Marketplace referral commission
Returns provision rate (0-1)Provision set aside for returns
Coupon (per unit)Coupon cost per unit

The per unit breakdown names the source of every line. A charge left out reads Not modelled.

Price point report

Select Price-point report to write the current price and cost scenarios to the project files as a cited profit model report.

07 Risk

Eight standing risks are listed, each stating what to check.

StatusMeaning
PendingNot yet looked at
Confirmed openThe risk is real and unresolved
MitigatedAddressed
AcceptedReal, and accepted as it stands

Record the evidence or the decision in the note field. An unconfirmed risk never blocks the conclusion. It caps the conclusion at Conditional Go and is named there.

08 Conclusion

The conclusion is bound by the opportunity score, the evidence confidence, the framework coverage and the risk sign off together. It recalculates as evidence arrives.

VerdictMeaning
GoNo unmet conditions
Conditional GoProceed, subject to the conditions listed
No-GoDo not proceed

Unmet conditions names every reason the verdict is not a clean Go. Select Recalculate after changing evidence, and Export decision memo to write the conclusion to the project files.