How Roarsome turns unanswered product questions into orders.
Roarsome went live with a Wassist agent on WhatsApp 15 days after the first call, just ahead of the five months that make up 70% of its year. In the 30 days to 22 October, shoppers who started a WhatsApp conversation spent 62% more per order than the store average, and those conversations influenced 1.6% of store revenue.
Average order value
+62%
when a shopper starts a WhatsApp conversation
30 days to 22 October 2026
- £176average order after a WhatsApp conversation, vs £108 across the store
- At least6.5%*of questions ended in an order, vs a 1.3% site conversion rate
- 1.6%*of store revenue (£1,408) influenced by WhatsApp conversations
- 57%of customer questions answered with no human involved
- 15 daysfrom first call to a live agent
* Influenced revenue uses a 30-day window, the same convention Klaviyo uses. It counts any order placed within 30 days of a conversation, so we show it as influence, not direct cause.
Roarsome skiwear for kids.
Roarsome makes character-led skiwear and outdoor clothing for kids: unicorn snowsuits, dino helmet covers, animal-themed layers. The brand started life as Dinoski (“roarsome skiwear for kids” was always the tagline) and is growing into a full kids' outdoor brand.
Where sales come from
- Shopify DTC 70%
- Retail 25%
- Clapham shop 5%
70% of turnover in five months
- J
- F
- M
- A
- M
- J
- J
- A
- S
- O
- N
- D
- Channels
- 70% direct-to-consumer on Shopify, 25% UK and international retail, 5% from its local shop on Northcote Road, Clapham
- Reach
- Distributors in the Middle East, Australia and Turkey, plus airport stores on the Gold Coast and in Sydney
- Scale
- Around 1,000 online orders a month at an average order value of about £110
- Seasonality
- 70% of annual turnover lands between September and January
- Customers
- Parents, who already run their lives through WhatsApp school and PTA group chats
Questions that wait don't convert.
Roarsome set itself a clear goal for the year. Moving to Shopify had already helped.
Parents buying kids' skiwear ask a lot before they commit: which size for a growing four-year-old, what the jacket is made of, whether it will arrive before the holiday, why a product says pre-order. When those answers don't come straight away, the shopper leaves, or comes back later, or doesn't.
Roarsome's goals for the year
Site conversion
1.3%to2%Average order value
£110to£150
Three things made that hard to fix:
One person held the inbox.
Every question went to Pippa, the store manager, who also runs the Clapham shop. Email and Gorgias tickets queued behind customers standing at the till.
Peak season.
70% of Roarsome’s turnover lands between September and January, which meant a spike in questions exactly when the team had least time to answer them.
The existing bot never got set up.
Roarsome’s helpdesk came with an AI bot, but nobody had time to train it.
“We were kind of left to our own devices. It was an action that fell down the list and has never really been done.”
There was also a reservation about the channel itself.
“WhatsApp is personal, and I don't know if I want to be contacted by a brand through WhatsApp.”
Ed said this on the first call. We agreed: Roarsome would use WhatsApp only for conversations customers start. No broadcasts, no newsletters.
An in-store manager for the digital store.
Wassist gave Roarsome an agent that does on WhatsApp what Pippa does on the shop floor: answers the question, suggests the right product, and brings in a person only when one is needed.
01
Answers product questions from the live catalogue.
Sizing, materials, stock, pre-order dates and delivery times come straight from Shopify. On Ed’s first test, the agent answered his question and recommended a unicorn snowsuit unprompted.
02
Sells.
The agent knows the current offers (3-for-2 on layers, spend more save more for Black Friday, a free beanie with every snowsuit) and sends product links with in-thread Shopify checkout.
03
Knows the customer.
Shoppers who log in link their Shopify account, so the agent can see past orders and suggest the next size up.
04
Turns questions into insight.
Every question is tagged by intent and product, so Roarsome can see what shoppers ask before they buy.
The handover
Hands over cleanly.
Three triggers send the chat to Gorgias:
- Faulty items
- Lost parcels
- “I want a person”
Pippa replies from Gorgias, and the customer sees it in the same WhatsApp thread.
Closing the ticket hands the chat back to the agent.
Live in 15 days, tuned every week.
Roarsome went from first call to a live agent in 15 days, then improved it in a 15-minute review every Friday.
10 August
Qualification
Will Chapman and Ed Lewis-Pratt shared the goal (1.3% to 2% conversion) and the bottleneck (one person answering every question).
14 August
The plan
Inbound only, customer-led. Start with a footer link to keep volume manageable, launch at about 80% quality, and tune from real conversations.
25 August
Live
The agent connected to Shopify and Gorgias, with handover rules for faulty items, lost parcels and “I want a person”. A test lost-parcel query reached Pippa, who resolved it in the WhatsApp thread.
Late August to September
Weekly tuning
Each Friday the team reviewed real conversations and adjusted the agent: duties on EU orders, pre-order products, rewards-point questions, size measurements from Pippa. Danielle and Pippa got their own logins to make edits directly.
September
Sitewide
The WhatsApp widget replaced the old site chat, and product pages got their own WhatsApp buttons.
October
Ready for peak
Black Friday offers went into the agent, and Roarsome began sharing question data with By Atlas to shape SEO and AI-search content.
Ed put the hands-on approach down as the difference.
You're keeping us accountable. There is a bulk of work that needs to be done to teach that agent. If you don't, it won't actually do what you need it to do.
Five times as likely to order. £68 more when they did.
In the 30 days to 22 October, a shopper who asked Roarsome a question on WhatsApp was five times as likely to order, and spent £68 more when they did.
Average order
Conversion
Shoppers who ask a question are already interested, so part of that gap is who chooses to chat. The point stands: the revenue sits in the questions people ask right before they buy, and answering them instantly is what Roarsome was missing.
Influenced revenue, share of store revenue*
1.2%
9 October
1.6%
22 October
Rising as peak season began.
Activity, 30 days to 22 October
| Activity | Count |
|---|---|
| Taps on the WhatsApp button | 215 |
| Customer questions handled | 123 |
| Answered by the agent alone | 70 |
| Handed to Pippa or Danielle | 53 |
| Product links sent in chat | 72, of which 18 were clicked |
| Customers who linked their Shopify account | 20 |
The powerful thing is it's not just customer service. It's turning it into a sales tool. You might have come in unhappy, but you're going to leave buying something.
Stuck on conversion? Read the questions.
If your conversion rate is stuck, look at the questions shoppers ask before they buy. At Roarsome, product and pre-purchase questions drove more influenced revenue than any other type of conversation. Five things made the difference.
- 1
Answer before they leave.
Sizing, materials and delivery dates are buying questions. An instant answer keeps the shopper in the moment they wanted to buy.
- 2
Let customers start the conversation.
Inbound only kept WhatsApp feeling personal, which was Roarsome’s biggest worry going in.
- 3
Start small, then go where the question happens.
Footer link first, then sitewide, then product pages, so the agent was tuned before volume arrived.
- 4
Tune it weekly with the people who know your customers.
Fifteen minutes on a Friday with the store manager beats a bot nobody has time to train.
- 5
Tell customers when a person will reply.
When the agent hands over, a clear response time keeps trust intact.
See what your shoppers are asking.
Install Wassist on Shopify to meet your agent, or book a demo and we'll walk you through Roarsome's setup.