A QR for the offline world (billboards, packs, flyers) and a link for online (WhatsApp, Instagram, SMS, ads). Both open the same AI chat, which qualifies people in English, Swahili or Sheng and sends only the ones who matter to your team.
It's also where you read the results. Ask your results questions and get answers back in seconds, export them to PowerPoint or CSV, and read the full AI chat behind every qualified lead.
Pilot on 3 placements · 6 weeks · live within 14 days
One chat, every entry point. Scan a QR offline or tap a link online, on any surface, in any industry.
Friday, 4:00 p.m. · your dashboard
Each QR and each link is unique to where you put it, so the chat knows exactly which billboard, pack or online post started it. By Friday you know where the next budget should go, offline and online.
Drop batch 4. Double Mombasa Road. That call used to be a feeling. Now it's a Friday habit.
From first scan to signed
Every stage below is a real event we count. You watch the crowd narrow from a scan to a signed customer, and you can see precisely which step loses people, so you fix that one step and leave the rest alone.
Surveys run as three stages (Scanned → Chatting → Completed) with no qualifying step. Pre-orders end in a paid reservation, deposit and all, right in the chat. Same words on your dashboard; there, every number is your own live count.
What you're actually buying
No new app for your customers to download. No scripted menu. Just a scan, or a tap, that turns into a real conversation and a clean handover to your team.
One chat behind every placement, offline and online.
A real conversation in English, Swahili or Sheng.
A clean handover: tagged, scored and delivered.
Monday, 8:40 a.m. · the team meeting
“Which of these contacts are worth a call?”
The weekend brought a pile of phone numbers. Somewhere inside is a handful of real prospects, and no way to tell which without calling every one.
A week
gone dialling a raw list to find the few who were ever going to buy
“What did that campaign actually do?”
Hundreds of thousands of shillings on boards, packaging and activations, and nobody can point to which one produced a single customer.
0
customers your last campaign can prove it produced
The loop runs both ways
Replies don't just leave for your inbox. Every lead, survey answer and pre-order also syncs into that campaign's AI workspace, so you can ask it questions in plain language and get answers drawn only from your own responses.
Every conversation's answers flow in, tagged by placement and ward.
Ask in plain language. No dashboard to learn, no analyst to wait for.
See which placements, wards and offers are winning, and which to cut.
Shift budget, tweak the offer. The next scans feed the same loop.
Every campaign gets sharper the longer it runs.
Turn any answer into a PowerPoint or PDF you can drop straight into a board meeting or client report.
Export your leads and responses to CSV in one click, for your own spreadsheets, CRM or BI tools.
Open the full AI chat behind any qualified lead and see exactly what they asked, and what they told you.
For agencies and product teams
The same analyst that answers inside Smart Campaign is available over an API, and it isn't limited to campaign data. Upload your own PDFs, contracts, price lists or exports and ask them questions in plain language, from your dashboard, your client portal or your field-force app. See what the API does →
# One key. It carries its own account, # so there's no tenant header to get wrong. curl https://api.smartcampaigns.ai/api/v1/\ conversations/$ID/messages \ -H "Authorization: Bearer rag_live_…" \ -d '{"content": "total connections by territory"}' # → exact figures, computed in the database { "assistant_message": { "content": "Entebbe led with 24, then Gayaza 21, Nakawa 16, Rubaga 14. Total 75 across 30 records." } } # X-RAG-Quota-Remaining: 4999
One key per client you serve, so usage and invoices stay separate. A key is bound to its own account. Send it from your server and it can only ever read that client's data, whatever else is in the request.
Push the rows you already export (form submissions, sales, listings) or the documents you already have: .pdf, .docx, .csv, .txt, .json, .md. They become a queryable workspace. Tabular data is computed from the structured records, so the arithmetic is exact.
Totals are exact and breakdowns add up to their own total. Where a figure can't be computed honestly, the API says so rather than returning a confident guess. Stream the answer or take it in one response.
Every billable response carries X-RAG-Quota-Remaining, and /usage/api_quota returns the same figures per key, so you always know what a client has used before the invoice arrives.
A question means a question asked. Reading answers back, listing workspaces or rating a reply are free and never count against your allowance. Create a key yourself under Settings → API keys; full reference and a live console at /api-docs, and the whole story for developers at /developers.
Do the maths on your own numbers
e.g. a Meta or Google lead-form contact
your billboard / pack / ad spend ÷ the scans it drives
platform usage: about KES 2.50–4 for a lead-gen chat, rising to ~KES 11 for a long survey
leads or chats your team closes one customer from
About KES 203,280 saved every month across 30 customers.
Start freeSelf-serve credit is consumed per conversation: around KES 2.50–4 for lead-gen and up to ~KES 11 for a long survey (that field is yours to change). You top it up like airtime, with no per-customer fee. “Cost to earn a scan” is your own media divided by the scans it drives, so both sides count the cost of reaching people. We assume it takes the same number of contacts to close a customer either way; once you're live, your dashboard reports your real numbers.
Fair questions
Not for nothing, no. But this QR answers the one question the person already has (the price, the details, the offer) instantly, instead of “call this number.” And if scanning isn't their thing, the exact same chat opens from a link you drop in WhatsApp, Instagram or an SMS, so you catch them offline and online. Because you see results per placement, you'll know within a week which ones earn their spot. Switch the demo above between Property, Retail and Events and time it yourself.
Scripted bots embarrass people. This one answers only from your documents: your price list, your menu, your brochures. You test it in a private sandbox with your own team before launch, it tells people it's an AI, and when it isn't sure it hands over to your people rather than making something up.
Only if you want one. You can buy conversation credit like airtime: top up, and it gets used when people actually chat, with nothing for procurement to approve. When your volume settles, an optional monthly plan buys the same credit at a better rate, and it cancels anytime with no contract. Or start with a pilot on three placements, at a fixed fee agreed up front.
If your campaign ends in a lead, a survey answer or a reservation, yes. The conversation is built from your own documents and your own qualifying questions, so a car dealership, a milk brand and a music festival each get a chat that sounds like itself.
You set hard caps before launch (per conversation, per phone, per day) and both of us get an alert if spending spikes. Your exposure is capped before the first scan, not after the bill.
Your key documents (price list, menu, brochures or survey questions; PDFs are fine), two or three places to put the QR, and one hour with us to set your qualifying rules. Beta campaigns go live within 14 days of kickoff.
Pricing
Someone scans and leaves? Free. The moment someone chats, you pay for that conversation from the credit you top up, whether or not they qualify. When a lead does qualify, it arrives fully scored, with the conversation, the placement it came from and a verified location attached. We price against what a qualified contact costs you today, not against what it costs us to run.
Doing nothing has a price. It just doesn't send an invoice.
Or self-serve
Monthly plans carry the best per-credit rate, and paying yearly gets you two months free. Prefer no plan at all? Top up any amount from $1.50 and the credit never expires.
KES 3,900.00
$30.00 USD / month
KES 39,000.00
$300.00 USD / year · KES 3,250.00/mo effective
KES 14,950.00
$115.00 USD / month
KES 149,500.00
$1150.00 USD / year · KES 12,458.33/mo effective
KES 37,050.00
$285.00 USD / month
KES 370,500.00
$2850.00 USD / year · KES 30,875.00/mo effective
Tell us about the placements you already run. If a pilot can be live within 14 days, we'll set it up with you, and you'll end it knowing which of them earns its keep.