# Josh Warwick

Josh Warwick builds Wassist, a service that uses AI agents on WhatsApp to help e-commerce brands guide customers through product discovery, checkout, support and returns.

- Profile: https://b2binfluencer.marketing/creators/josh-warwick/
- LinkedIn: https://www.linkedin.com/in/joshwarwick
- LinkedIn followers: 4,324
- Topics: artificial-intelligence, ecommerce, ecommerce-retention
- Based on 13 dated posts, latest 2026-10-08

## Summary

Josh Warwick is most useful for explaining AI agents on WhatsApp for e-commerce discovery, checkout, support and returns.

- Best for: A conversational commerce platform or beauty, fashion or lifestyle brand with repeat purchase and service-heavy operations.
- Not for: A paid placement or sponsored post judged on reach, follower demographics and past campaign performance.
- Check first: Ask for commercial format, rate card, deliverables and audience analytics before contacting.

### Takeaways

- Josh Warwick says WhatsApp had 86% month-one retention versus Instagram at 78%.
- A WhatsApp exchange with a Prada store became the origin story for Wassist.
- Josh Warwick says map the whole journey before deciding what the product is.

## Main ideas

- **Conversation is replacing navigation as the shopping interface**: Warwick argues that websites dominated for 30 years because they mirrored desktop computing — menus, windows, a mouse — and that mobile changed the screen and added touch without changing the core interaction. Large language models change that: a conversation can handle product discovery, recommendations, support and returns more efficiently than site navigation once the model understands intent and order history.
- **Pick the venue by retention, not by fashion**: If conversation is the interface, Warwick asks where the conversations are and points to attention data: two billion people open WhatsApp daily, and its reported month-one retention and daily-to-monthly ratio put it ahead of the apps brands usually chase. They say first-principles reasoning led to WhatsApp while building Wassist, and the statistics validated the choice afterwards.
- **WhatsApp commerce is already proven — outside EMEA the constraint was staff, not demand**: Warwick argues that messaging a brand to buy, ask questions and resolve problems is already the default in Brazil, India, Mexico and Singapore. In EMEA, the barrier was scale: a conversation required a staff member to answer, which only luxury brands could afford. LLM-based agents break that barrier because labour costs no longer rise with every conversation, leaving support and success teams to handle the difficult cases.
- **Give every brand a sales associate at any price point**: The founding idea is that a good conversational sales experience existed only at luxury price points because a human delivered it, and that agents can make it affordable elsewhere. Warwick treats the personal retail experience as the product specification, not a marketing anecdote.
- **Map the whole journey before deciding what the product is**: Warwick’s working priority is to define Wassist by the customer’s actual path, not by a software category. Two weeks after launch, the team stopped describing it as a customer service or marketing tool because onboarded brands were using it across more of the buying journey than those labels suggested.
- **Work in public, and convene the people doing the work**: Warwick regularly shows the process, not just the conclusion. They publish weekly operating snapshots — sales onboarding, integrations, events, judging a hackathon and hosting a brand dinner — under the framing that early-stage work means wearing several hats at once. The same instinct drives them to organise practitioner gatherings, not just publish arguments.

## Work to read

- [Origin story: how a WhatsApp message to Prada became Wassist](https://www.linkedin.com/posts/joshwarwick_ive-been-sharing-the-origin-story-of-the-activity-7497656341801631744-NCNZ)
- [Why WhatsApp, not a website, becomes the default commerce channel](https://www.linkedin.com/posts/joshwarwick_whatsapp-is-about-to-become-the-default-commerce-activity-7462071748566618112-ty6A)
- [The retention numbers behind the WhatsApp bet](https://www.linkedin.com/posts/joshwarwick_whatsapp-ecommerce-buildinpublic-activity-7477638182222667776-LLE8)
- [Two weeks after launch: how the product description changed](https://www.linkedin.com/posts/joshwarwick_buildinpublic-activity-7474760560430161921-zMyO)
- [Running a first event: a panel on scaling e-commerce into 2027](https://www.linkedin.com/posts/joshwarwick_were-opening-up-invites-to-our-first-event-activity-7498676701816811520-4mdw)
- [A week of wearing every hat](https://www.linkedin.com/posts/joshwarwick_whats-on-this-week-in-the-early-stages-activity-7479818376157016064-uOsL)
- [Shortlisted to speak at E-commerce Berlin Expo](https://www.linkedin.com/posts/joshwarwick_ive-been-shortlisted-to-speak-at-e-commerce-activity-7510252453762596864-uH96)

## Brand relationships

- **Wassist** (owned_business)
- **Prada** (mention)
- **Balderton** (mention)
- **PerfectTed** (mention)
- **Willy Chavarría** (mention)
- **eCom Collab Club** (mention)
- **Cursor** (mention)
- **WIRO** (mention)
- **E-commerce Berlin Expo** (mention)

## Partnerships

Warwick's posts don't offer a partnership or sponsorship format to buy. They focus on building Wassist, the AI-agent service that runs e-commerce conversations on WhatsApp, and on events and forums where that work gets discussed: a panel event they are opening invites to at Balderton, covering what worked for scaling e-commerce in 2026 and plans for 2027; a brand dinner hosted with Tom Rees and WIRO; a shortlisted speaking slot at E-commerce Berlin Expo that depended on a public vote; and a working week of sales onboarding, integrations and hackathon judging. For a buyer, that points to a speaking or advisory engagement — explaining LLM-driven customer journeys on WhatsApp — rather than a paid content placement or an affiliate deal. The posts include no rate card, deliverable list, sponsorship policy, media kit, booking portal or partnership inquiry link. Approach them through LinkedIn to ask about scope, fee, and whether they would appear as themselves or as Wassist.

## Alternatives

- Ishay Tentser: Choose Ishay Tentser when the brand has already accepted messaging as a channel and the live problem is operational: WhatsApp conversations sitting outside the CRM, customer relationships tied to whichever number a rep happens to own, or handoffs that depend on a colleague forwarding messages. The two posts read as an operational checklist rather than a market argument. One offers a diagnostic a support or sales lead can answer directly: does a new WhatsApp chat wait for someone to create the contact record? Does context stay split across individual numbers? Does every follow-up require a manual hop between WhatsApp and the CRM? It then lists the fixes: confirm the CRM integration logs messages and creates the records the workflow needs; bring connected numbers into one shared inbox with visible assignment; trigger WhatsApp messages from CRM events where the integration supports that; and let AI list chats, read history, draft replies, label, assign and triage across the workspace. The same post names the native CRM connections it claims, including HubSpot, Pipedrive, monday.com, Zoho, Salesforce, Close CRM, Bitrix24, SugarCRM, GoHighLevel, Copper, Teamleader Focus and Nutshell. The other post covers ownership: what happens when a rep leaves and takes a personally owned number, why a company number keeps thread history and offboarding under the business, and how to migrate in practice by running both numbers, telling customers where the rep is reachable, then retiring the old setup. For an advisory engagement focused on integration sequencing inside an existing sales organisation, this is closer to the work than a talk about why conversation beats a website.
- Alex Fedotoff: Choose Alex Fedotoff when the real question is whether a channel or market works economically, rather than how the buying journey should be shaped. One post walks through supplement unit economics with named benchmarks — $89 average customer acquisition cost against an average order value between roughly $45 and $90 — and shows why a brand at those numbers loses money on every first order and is betting the business on retention: subscription rate above 40%, lifetime value above $200, refunds below 3%. It then works through a concrete example: a roughly $6M run-rate supplement brand at $78 front-end acquisition cost and $54 order value that believed it was profitable for eight months while losing about $11 per first order, based on a 38% subscription assumption against an actual 27% rate and a $124 lifetime value. It contrasts supplements with fashion at $37, pet products at $23 and food at $51. The second post makes the case for expanding into India with figures a planning conversation can test: 1.4 billion people, Meta CPMs of $0.50 to $2 across many health and beauty niches, a $147 billion e-commerce market growing 18–20% a year, a D2C market projected at $60 billion by 2027, 270 million online shoppers, UPI handling about half of e-commerce payments, and a reported US skincare launch where the same landing page and ad structure moved acquisition cost from $44 in the US to $6 in India at a $29 price. If the question is where to put next year's budget and whether payback holds, this is a sharper contribution than an interface argument.
- Nazmul Hassan joy: Choose Nazmul Hassan joy when the work is execution on the current web-and-paid-social model: building a Shopify storefront or restructuring a Meta ad account. One post lays out an account-structure position a media buyer can apply directly: separate proven winners from testing so unproven ads do not consume budget that already works; prefer broad targeting when the creative clearly conveys the customer, product, lifestyle and problem instead of building many small interest audiences; treat creative diversity as the audience strategy; give retargeting its own role rather than mixing it with cold acquisition; and scale only when the account has earned it. It also reports results of $19,391.77 in spend producing $225,526.60 in purchase value from 1,071 purchases, a 13.21x average purchase ROAS and individual campaigns between 10.38x and 17.29x. The other post offers store-design guidance for a women's fashion Shopify store: use a clean layout with high-quality model photography and matched typography; customise homepage sections, navigation, collection pages and product pages rather than accepting theme defaults; organise collections to keep discovery simple; put sizing, fabric, care and shipping details on the product page; and check every image, menu, variant and popup at mobile sizes. For a brand constrained by its storefront and ad account, this is closer to the deliverable than a talk.

---
AI-assisted summary of public posts. Audience demographics are not available. Engagement appears only where we checked recent posts. Method: https://b2binfluencer.marketing/methodology/