An electrical contractor with six employees handled dozens of active job sites each year. The professional side worked well. The administrative side ran through WhatsApp: every employee who bought materials photographed the invoices and sent them to the owner, who passed a pile of photos to the accountant at the end of the month.
Starting point: the month closed when the last invoice was found
Our discovery process revealed the following:
- Between 80 and 120 supplier invoices per month. Most were photographed on a phone, some arrived by email, and others stayed in a workwear pocket.
- Every month, the accountant returned a list of missing documents, and the owner began searching through WhatsApp, email, and the vehicle.
- Average month-end close: 12 days after the end of the month. Tax advances were paid based on estimates, and the real picture of project profitability emerged, at best, two months late.
- According to the owner, nearly six hours a week went into photographing, forwarding, finding, and filing documents.
This pattern repeats across field-service businesses worldwide. Industry surveys put the cost of manually processing an invoice at more than ₪50, while automatic capture removes the vast majority of that work.
What we built
Together with the business’s accountant, who was a full partner in the process, we set up one document pipeline:
- One intake channel for every document. Employees still take photos in WhatsApp, but they now send them to a dedicated business number instead of the owner. Invoices that arrive by email are routed automatically from a dedicated inbox.
- Automatic recognition and capture in iCount. Every image goes through text recognition. The supplier, amount, invoice number, and date are extracted automatically, and the document is stored in an organized, readable form. Duplicates, such as the same invoice photographed by two employees, are detected and blocked.
- Project assignment with one tap. The employee receives one WhatsApp question: which job site does this purchase belong to? The answer tags the invoice and builds a true cost picture for each project.
- Missing details handled in real time, not at month-end. An invoice with a missing detail prompts the employee on the same day. The accountant receives direct access to the system and retrieves the material in order.
Results after two months
- Month-end close fell from 12 days to four.
- Weekly paperwork time fell from six hours to less than forty minutes, mostly spent on one-tap approvals.
- Zero lost invoices during the two measured months, compared with an average of seven missing invoices in each previous month.
- For the first time, the owner can see material costs for each job site while the project is underway rather than two months after it ends. One project showed a material-cost overrun during the work, and pricing for the next stage was adjusted accordingly.
The accountant put it best: “Instead of chasing documents, we are finally talking about the numbers themselves.”
What this story teaches
Paperwork is not a discipline problem. It is a process problem. When the easiest way to handle an invoice is also the correct way, one WhatsApp photo, the problem largely disappears.
Questions we are asked about this kind of process
We already have accounting software. Do we have to move to iCount? No. The solution is built around the tools already in place. We have worked the same way with Green Invoice, Rivhit, and Hashavshevet. The choice here was made together with the business’s accountant.
Does automatic recognition ever make mistakes? Sometimes, especially with poor-quality photos. That is why any document the system is unsure about goes to a brief human review rather than being captured incorrectly. For this client, the vast majority of documents were processed without manual intervention.
Our financial information is sensitive. Where is it stored? In the secure accounting system already used by the business and its accountant, not with us. We build the pipeline; we do not hold the information.
The client’s identifying details were removed at their request. The data was measured with the client and accountant before and after the project.
