TL;DR
- Before the marketplace can automate liquidity, the team needs to understand how liquidity forms manually.
- Manual operations should be structured enough to generate data: who matched, who requested, who declined, and why.
- The goal is not to stay manual forever. The goal is to learn which manual steps are worth productizing.
The first marketplace exists as an operating workflow
A marketplace does not become useful the day the catalog goes live. It becomes useful when a buyer can find a relevant brand, trust the information, take an action, and get a timely response. Early on, much of that response can be manual.
That is not a weakness. It is the correct way to learn. Manual workflows reveal which categories have demand, which buyer questions repeat, which brands are hard to operationalize, and which parts of the product need automation first.
The core manual objects
- Brand card: company, category, country, pricing readiness, imagery readiness, inventory mode, buyer controls, notes.
- Buyer profile: buyer type, country, categories, price points, delivery windows, open-to-buy timing, brands carried, notes.
- Category interest: supply, demand, and rep coverage by category and country.
- Manual match: buyer, brand, category, country, status, notes, outcome.
- Preview event: category/country theme, invited brands, invited buyers, RSVPs, requests, feedback.
These objects do not need to be perfect. They need to be consistent enough that the team can learn from them.
What the team should do after Step 1
- Create or update the contact and company record.
- Capture source, consent, referral, and audience type.
- Create an application record with status 'Step 1 complete.'
- Ask for Step 2 profile details based on audience.
- Queue the contact for the right nurture sequence.
- Log category and country interest even if the profile is incomplete.
What the team should do after Step 2
- Calculate the audience-specific score.
- Assign a score band and status.
- Update category/country density.
- Push the record to CRM or keep it queued if the integration is not enabled.
- Send an internal alert for high-score leads.
- Route strong leads into review, interview, preview event, or expansion waitlist.
- Track what happened next so the score can be improved.
Manual matching should be opinionated
A manual match should not mean 'these two records both mentioned womenswear.' It should mean the buyer's category, price point, delivery window, country, and store profile plausibly fit the brand's assortment and operating promise.
The team should log why a match exists and what happened after the intro. A request, sample, save, rejection, or silence all teach the marketplace something.
- Good match: buyer and brand align on category, price, timing, geography, and readiness.
- Weak match: category aligns but price, timing, or operational requirements do not.
- Learning match: intentionally tested to learn a segment, but not treated as proven demand.
- Rejected match: buyer feedback should be captured and reused.
How manual work becomes product roadmap
The product roadmap should follow the manual work that repeats. If the team repeatedly asks brands for missing size/color data, build a readiness checklist. If buyers repeatedly request ready-to-ship filters, prioritize ATS visibility. If reps need attribution on every intro, build the attribution model before public scale.
Manual operations are not separate from product. They are the discovery process for the product.
The prelaunch operating layer is valuable because it lets Apparel Market act like a marketplace before everything is automated. The discipline is to capture every manual action as data.