Preboot for e-commerce
An AI that sells from your live catalog at 3 AM. Journeys that collect the revenue you already earned — the cart, the COD order, the second purchase. And Loop behind them, watching campaigns, couriers, and margin, arriving with decisions instead of dashboards. Approval-gated, reversible, on the record.
Sound familiar?
Most of my carts just… vanish.
Published averages put cart abandonment around 70% across e-commerce — and most stores never say a word to those buyers again.
COD returns are eating my margin.
You pay the courier both ways to find out the customer changed their mind at the door.
I run the store on numbers I can't see.
Revenue per campaign message. Refusal rate per courier, per city. Repeat rate by first product bought. Most stores have never seen these numbers once.
None of this is a staffing problem. It's a context problem — and an AI with full context solves it.
The funnel
Three journeys that chase the money the store already made — the abandoned cart, the unconfirmed COD order, and the second purchase.
Cart recovery built from a sentence — and rewritten the week its conversion decays, not the quarter after.
Every abandoned cart followed up 24/7 — designed to recover 10–20% of the ~70% that walk away.
Test new recovery offers with preview and one-click rollback — a failed test costs minutes, not the funnel.
COD orders confirmed in chat before dispatch — refusals caught while they still cost nothing, designed to cut RTO 30–50%.
The confirm flow built from a plain-language description — tuned by city or order value, without a developer.
Unconfirmed orders held, confirmed ones pushed to the courier automatically — no copy-paste dispatch.
Post-purchase journeys — check-in, cross-sell, replenish — designed to lift repeat purchase 15–30%.
Describe the journey per product line — Loop assembles it in minutes, explains branches, fixes laggards.
'First-time buyers, 30 days silent' stays fresh automatically — the journey always targets the right people.
Loop, watching
Loop watches campaigns, revenue, and regulars — and arrives with the call already made: reasoning, expected outcome, and risk attached.
'Kill it, keep it, or rework it' — with expected outcome and risk attached. Decided in minutes, not at month-end.
Loop monitors every live campaign and proposes stopping losers — approval-gated, designed to cut waste 20–35%.
Separates a genuinely fatigued campaign from one bad week — before you kill a winner by mistake.
A dip traced to its real cause — stockout, courier delays, or ad fatigue — before money moves on a hunch.
Orders, chats, campaigns, and events already live in one queryable warehouse — the answer exists on day one.
Follow-up questions answered as charts in chat — a data team's week of digging, without the data team.
Regulars going quiet flagged in days — unprompted — instead of surfacing in next quarter's revenue.
'Drifting regulars' is a living segment mined from behavior — always current, never a stale export.
Win-backs built on each group's actual bestsellers — designed to lift reactivation 2–3× over a generic blast.
The front line
The AI sells from the live catalog and clears the question that buries every store inbox — your team keeps only the conversations that need a human.
Sells from live stock and the buyer's whole history — objections answered, order closed in chat, 24/7.
Understands the photo a customer sends, matches the product, and answers in their language.
Live prices and stock from Salla and Shopify — the bot never promises what the shelf can't.
The AI answers with the actual order status and full history — WISMO handled instantly, designed to cut the human load 40–60%.
Live tracking from Salla, Shopify, and couriers inside the chat — customers stop needing to ask twice.
Validated publish and human handoff with reasons — the bot never invents a delivery date.
Run the store on facts
Couriers, agents, and the weekly review — the management layer no chatbot ever touched.
Loop reads courier tracking through custom integrations and proposes re-routing weak lanes — on your approval, designed to cut RTO 20–35%.
Late deliveries and refusals attributed per courier, per city — the real driver, not an anecdote.
A re-routing brief with cost, expected refusal drop, and risk — courier choice becomes a decision.
Reply time, resolution, and converted chats per agent — from real conversations, so reviews argue facts, not memories.
Targets set from actual agent behavior, not gut feel — incentives your team reads as fair.
The weekly review lands every Monday at 7 AM — busy season, peak weeks, vacations — it cannot forget or lapse.
Sales, AOV, refusals, campaign ROI — assembled from the warehouse, charted, delivered before coffee.
It leads with what changed and what you've been ignoring for weeks — not a wall of numbers.
Before & after
The remaining six moments from our store map — run by hand, and run on Preboot.
Back-in-stock demand
A restock nobody hears about — the demand lost silently
Waitlist joins in chat; the alert fires the hour stock syncs — designed to recapture 20–40% of stockout demand
The campaign send
A blast to the full list, risking the number
Warm-up, quality watch, segment-first sending — designed to lift revenue per send 25–50%, with the number protected
Campaign ROI
Guessed from clicks and a feeling
Campaign→conversation→order captured natively — 'which message made money' becomes a fact, not a model
Demand signals
Show up weeks later, in the P&L
Every unanswered 'do you have…?' counted across the whole inbox — demand your ads never had to find
What to stock next
Gut-feel launches
Inventory and offers aligned with what the market keeps asking for — hit rate up, dead stock down
Rising CAC
Another month of spend on a hunch
Decomposed — creative, audience, or price — with a budget-shift brief attached, designed to cut blended CAC 10–25%
"Published averages" are industry benchmarks from public studies, not Preboot results. Your own numbers will live on your dashboard — that's rather the point.
Done right by default
Your WhatsApp number, protected.
Warm-up, quality monitoring, and segment-first sending. Blasting lists is how stores lose their number; Preboot is built so you never blast.
Opt-ins and opt-outs, automatic.
Consent is tracked per customer and honored by every campaign and journey — without you thinking about it.
Every AI conversation, auditable.
Session summaries, handoff reasons, reasoning traces. You can read exactly what the AI said, did, and why.
Nothing ships without you.
Loop's changes follow the same rules as yours: previewed, approved by you, reversible in one click.
Yes — both connect out of the box. Orders, carts, and customers sync automatically, and the chatbot answers from your live catalog. Other systems connect through custom integrations.
It answers from your synced catalog, not from imagination. Changes to the bot are validated before publish, and when it isn't sure, it hands the conversation to your team with the reason attached.
Yes. Every change — yours or Loop's — is previewed before it goes live and can be rolled back in one click. That includes offers, cart recovery, and the chatbot itself.
Protecting your number is a core design goal: warm-up, quality monitoring, and segment-based sending instead of blasts — blasts are the single biggest cause of blocks.
It measures what's already on the record: first reply times, resolutions, and converted chats from real conversations. Managers see it under roles and permissions, and targets built from it tend to be accepted precisely because they're factual.
No. Loop proposes — stop this campaign, re-route this courier lane, shift this budget — with the reasoning, expected outcome, and risk attached. Nothing runs until you approve, and every change rolls back in one click.
Message us on WhatsApp — our own number runs on Preboot, so the first thing you'll see is the experience your customers would get.