Giving Electronics a Second Life

In an earlier Founder Diary, I wrote about my friend Mubeen and a challenge raised by one of his clients: an electronics refurbishment company in Birmingham, UK.

At the time, I only understood the problem at a high level.

Revisiting a problem I only half understood

Recently, we had another review with the client and looked more closely at their spreadsheets, operating data, and day-to-day workflow. That discussion helped me understand why this is not simply an inventory problem.

It is a problem of translating a complicated physical process into reliable financial information.

The company purchases used electronics in bulk, often as mixed shipments. The goods may be purchased by batch, container, or total weight, rather than as clearly identified individual products.

Buying potential, not finished inventory

When a shipment arrives, the team does not yet know its true value.

Each item must be inspected, sorted, tested, and graded. Some products can be cleaned and resold. Some require repairs. Some are dismantled for reusable parts. Others can only be sold as scrap.

The company does not buy finished inventory. It buys potential value.

That value is gradually discovered through the work of the team.

When cost can't follow the product

This creates a difficult costing problem. The company must allocate the original purchase price, shipping, labor, repair parts, storage, and other shared costs across products that may follow completely different paths.

One batch can eventually become refurbished devices, spare parts, and recyclable materials.

The revenue may be visible, but the true cost — and therefore the real profit — is much harder to calculate.

When cost cannot follow the physical product, profit becomes an estimate.

The limits of Excel

The company has developed detailed Excel spreadsheets to manage this process. Those spreadsheets contain a great deal of operational knowledge and reflect years of experience.

But they also show the limits of Excel.

Excel can record what happened. It does not naturally understand how one shipment was split, transformed, repaired, dismantled, and eventually sold through several different channels.

Traditional accounting software usually sees only the original purchase and the final sale. A full ERP may be able to model the process, but it is often too expensive, too rigid, and too difficult for a small team to maintain.

Where AI makes a difference

This is where I believe AI can make a meaningful difference.

In the past, software for a specialized workflow like this would require extensive customization. Every product type, grading method, repair rule, and cost-allocation process would need to be designed in advance.

AI allows the system to learn from the information the business already produces. It can read supplier invoices, shipment lists, spreadsheets, repair notes, and product photos.

Specifically, it can:

• Classify items, identify condition grades, summarize repair work, and detect missing information
• Suggest how shared costs should be allocated based on actual outcomes and historical recovery rates
• Identify unusual margins, inconsistent quantities, and batches performing differently from expectations
• Reduce the amount of administration required to understand the business

The warehouse team should not need to become accountants or ERP specialists.

Describe the reality, let the system handle the accounting

They should record what happened in the real world: an item was received, inspected, repaired, dismantled, sold, or recycled.

The system should translate those actions into inventory movements, work in progress, cost allocation, and financial results.

This reflects one of the values we are building into SnapLedger:

The user describes the business reality. The system handles the accounting complexity.

AI does not replace accounting rules, structured data, or human judgment. But it can significantly reduce the effort required to convert messy operational information into accurate and useful financial records.

A problem much broader than one company

That is particularly important for SMEs.

Many small businesses do not have simple operations. They simply have fewer people managing the complexity.

They may still rely on Excel because traditional systems are too expensive, too difficult to implement, or too disconnected from the way their business actually works.

The Birmingham refurbishment company is one example, but the underlying problem is much broader. It exists wherever business owners and employees spend large amounts of time maintaining spreadsheets, reconciling different files, manually tracking projects, and trying to understand why their financial results do not match operational reality.

If your company is still using Excel for complicated management and record-keeping, struggling to calculate accurate costs or profits, or finding that project management requires too much manual work, I would be very interested to hear from you.

Please write to me at richard.tang@snap-ledger.com.

Your experience may help shape what SnapLedger builds next.