Problem → service matcher
Choose a problem, size and budget and press Match. Without JavaScript, use the full matrix below — it contains every recommendation this tool can make.
Every problem × budget band
| Problem | A — under US$500 one-off or under US$50/month | B — US$500 to US$5,000 | C — US$5,000 to US$25,000 | D — over US$25,000 | Note |
|---|---|---|---|---|---|
| P-01 Customer service | Tool + S-05 | S-02 + S-05 | S-03 + S-04 + S-06 | S-01 + S-03 + S-04 + S-06 | WhatsApp first. Under band A a listed platform with a chatbot builder covers FAQs; anything that must read your stock or bookings needs S-04. |
| P-02 Lead generation | Tool + S-05 | S-02 + S-04 | S-03 + S-04 + S-06 | S-01 + S-03 + S-04 + S-06 | Capture and qualification only work if leads land in one place; budget for the CRM connection before the AI. |
| P-03 Marketing | Tool + S-05 | S-05 + S-02 | S-03 + S-05 | S-01 + S-03 + S-05 + S-06 | Most marketing wins are training plus a content workflow, not a build. Beware vendors selling ads management as AI. |
| P-04 Sales | Tool + S-05 | S-04 + S-02 | S-03 + S-04 + S-06 | S-01 + S-03 + S-04 + S-06 | Quote generation and follow-up sequences need your price list and CRM connected; that is integration work. |
| P-05 Operations | S-05 + Tool | S-01 + S-02 | S-03 + S-04 + S-06 | S-01 + S-03 + S-04 + S-06 | Operations problems are usually data problems. An audit before a build saves the build. |
| P-06 Finance | Tool + S-05 | S-02 + S-04 | S-03 + S-04 + S-06 | S-01 + S-03 + S-04 + S-06 | Any invoice automation must respect ZIMRA fiscalisation (FDMS). Ask how the vendor handles fiscal invoices before anything else. |
| P-07 Administration | Tool + S-05 | S-05 + S-02 | S-03 + S-04 | S-01 + S-03 + S-06 | Inbox, forms and filing are the cheapest wins; training on existing tools often beats a build. |
| P-08 HR and recruitment | Tool + S-05 | S-02 + S-05 | S-03 + S-04 + S-06 | S-01 + S-03 + S-04 + S-06 | CV screening touches personal data of applicants: your data-controller licence tier (SI 155 of 2024) may change. |
| P-09 Reporting | S-05 + Tool | S-01 + S-02 | S-04 + S-03 + S-06 | S-01 + S-04 + S-03 + S-06 | If the numbers live in five spreadsheets, the AI will summarise five disagreements. Fix the source of truth first. |
| P-10 Document processing | Tool + S-05 | S-02 + S-04 | S-03 + S-04 + S-06 | S-01 + S-03 + S-04 + S-06 | Extraction accuracy depends on your scan quality. Pilot on 200 real documents before you sign a build. |
| P-11 Voice | S-05 | S-02 | S-02 + S-03 + S-06 | S-01 + S-03 + S-04 + S-06 | English voice agents work today; Shona and Ndebele recognition is still maturing. Always pilot with real callers first. |
| P-12 Computer vision | S-05 | S-01 | S-02 + S-03 | S-01 + S-02 + S-03 + S-04 + S-06 | Cameras, lighting and power come before models. Under band B the honest answer is an audit, not a build. |
Codes: Tool = off-the-shelf subscription (see providers); S-01 audit; S-02 pilot; S-03 implementation; S-04 integration; S-05 training; S-06 managed support. Size adjusts the advice, not the shapes: Micro (0–5 staff): At this size, buy a tool and learn it; a custom build rarely pays back. Keep one owner for the tool. · Small (6–30 staff): Name one internal owner and one process. Pilot before build; insist on a fixed-scope quote. · Medium (31–75 staff): You likely hold personal data on more than a thousand people: check your SI 155 licence tier and name the data owner in the contract. · Large (76+ staff): Run an audit first, procure with the 60-question bank, and separate implementation from managed support in the contract.