Spur

Spur: One Inbox For WhatsApp, Instagram And The Website

The positioning is direct: sell more, support better, automate everything — a multi-channel AI agent for marketing and customer support.

The channel list is the thing to notice, because it describes where commerce conversations actually happen rather than where a software company would prefer them to.

For a very large share of online retail, the buying conversation is a WhatsApp message asking whether something is in stock, an Instagram DM asking about sizing, or a comment under a post. The website chat widget is often the least used of the four.

Those channels have historically been served by separate tools, separate logins and separate people, which is why a customer can get three different answers depending on where they asked.

One inbox with built-in ticketing, custom fields and multi-agent collaboration is the unglamorous fix, and it matters more than the automation layered on top. The site's own before-and-after framing names the real symptom accurately: chained to your inbox day and night, leads slipping away off-hours.

The buying conversation is a WhatsApp message about stock or an Instagram DM about sizing. The website chat widget is often the least used channel of the four.

Comment To DM And Where Leads Leak

Comment To DM is described as automating replies to comments — sending a DM to share a link, qualify a lead, or ask someone to follow if they are not already.

It is a small feature addressing a genuinely large leak.

A post that performs well collects hundreds of comments asking price, availability and where to buy. Every one of those is a person who has raised their hand, and answering them individually is impossible at volume — so most go unanswered, and the interest evaporates within the hour it took to lose it.

Moving the conversation from a public comment into a direct message also changes what is possible: you can send a link, ask a qualifying question and take an order in a DM, none of which works in a comment thread.

The judgement required is about tone. An automated DM that reads as a broadcast performs worse than no reply at all, and asking someone to follow before you have helped them inverts the exchange. The mechanism is sound; the copy decides whether it lands as service or as spam. Website-focused assistants like Chatling solve the same qualification problem on a single channel.

What Does An Actionable Agent Change?

The first listed capability is actionable AI: integrating with external systems to pull real-time data, update orders or book meetings automatically.

That distinction — between an agent that answers and an agent that acts — is the whole difference between deflecting a ticket and resolving one.

A chatbot that can only answer questions handles the easy half of support. Where is my order is a question about live data, and without a systems integration the best possible answer is a link to a tracking page the customer has already checked.

Change my delivery address, cancel this item, apply this discount — these are the requests that actually consume support time, and they all require the agent to do something rather than say something.

The integration list reflects the commerce focus: Shopify and WooCommerce for stores, Razorpay and Stripe for payments, Shiprocket and Return Prime for logistics. That combination is what makes actionable claims credible, because an agent can only update an order in a system it is connected to.

The stated outcomes — 90 percent fewer escalations, ten times faster resolutions, 30 percent higher CSAT — carry no methodology, and should be read as claims rather than benchmarks. Tidio Lyro makes the same argument on website chat, where the agent's actions come from helpdesk integrations rather than commerce ones.

• Connect the order and logistics systems first — that is where resolution comes from.
• Write automated DM copy that reads as help, not as a broadcast.
• Set escalation rules before launch, not after the first complaint.
• Check WhatsApp Business API rules for your message types.
• Watch what the agent handles badly, weekly, in the shared inbox.

Handover, Escalation And Knowing The Limit

Seamless handover is listed as a headline capability: when complex issues arise, the AI passes conversations to human agents.

This is the feature that determines whether automation improves customer experience or destroys it, and it is worth more scrutiny than it usually gets.

Every customer has been trapped in a loop with a bot that could not help and would not escalate. The damage is not the unanswered question; it is the message the experience sends — that the company would rather not hear from you.

A customer who reaches a person after two exchanges has had a good experience. The same customer after ten has decided something about the brand that no resolution reverses.

The design questions are therefore concrete: what triggers a handover, how quickly, whether the customer can request one at any point, and what happens outside staffed hours. None of that is answered on a marketing page, and all of it should be established in a trial with real conversations rather than a demo.

Getting it right is also what makes the automation defensible internally — support teams resist these tools mainly because they have watched one make their job harder.

A Negative Review Left On The Page

Among the five-star testimonials on Spur's homepage sits a one-star review, published in full: too complicated to set up, could not receive a one-time passcode, and the reviewer states plainly that they did not understand it.

Underneath it is a response inviting anyone facing setup difficulty to get in touch.

We have read several hundred vendor sites for this programme and this is the first time we have seen a company leave a negative review on its own homepage. It deserves saying clearly, because the alternative is universal and unremarkable: curate the wall, show only the fives, and let the buyer discover the friction after purchase.

It is also useful information rather than merely admirable. The complaint names a real and specific onboarding problem — WhatsApp Business API verification involves phone verification, Meta approval and business documentation, and it genuinely does defeat people.

A buyer reading that review learns something true about the product that no feature list would have told them: budget time for setup, and expect the channel connection to be the hard part rather than the automation. Enterprise support platforms such as Forethought face the same onboarding reality with more implementation support attached.

Automating Support Without Losing Customers

Automation goes wrong in one direction only: the customer who needed a person and could not reach one.

This order protects against that.

1. Budget real time for channel verification — it is the hardest part of setup.
2. Connect order and logistics systems before switching the agent on.
3. Load knowledge sources from your actual support history, not your FAQ page.
4. Define escalation triggers and let customers request a human at any point.
5. Launch on one channel and one intent, then widen.
6. Read the conversations the agent handled badly every week.

Spur: What Runs On Which Channel

• Four channels, one inbox — WhatsApp, Instagram, Facebook and website chat with shared ticketing.
• Actionable agents that pull live data, update orders and book meetings.
• Comment To DM, moving public interest into a private conversation automatically.
• Campaigns and abandoned cart flows with a stated 80 percent-plus delivery rate.
• Meta Business Partner and Shopify Partner, GDPR compliant, with a published DPA.

Spur: Channels And Compliance