Decision Intelligence: Where SMEs Are Losing Money Without Knowing It

Your ERP knows. You don't.

Somewhere in your accounting software, there's a supplier who has quietly raised prices 11% over the past six months. Nobody flagged it because nobody asked. Your ERP recorded every invoice. Your team reconciled every payment. The data was there the whole time.

That's the pattern we see over and over in Italian manufacturing SMEs: the information exists, scattered across five or six systems, but no one connects the dots until the damage is already in the quarterly numbers.

A purchasing manager at a EUR 25M automotive parts company told us he discovered a EUR 180K margin leak on a single product line — six months after it started. The data was in the ERP. The price change was in the supplier emails. The margin erosion was in the accounting system. Three systems, one problem, zero alerts.

Business intelligence shows you the past. That's the problem.

If you have Power BI dashboards or Excel reports from your finance team, you already have business intelligence. It answers the questions you know to ask: last quarter's revenue, margin by product line, days sales outstanding.

What it doesn't do is tell you what you don't know.

	Business Intelligence	Decision Intelligence
Question it answers	"What happened last quarter?"	"What's happening now that you should act on?"
Data approach	Single source, manual consolidation	Cross-source, automatic connection
Output	Charts and reports to interpret	Specific recommended actions
Who does the work	Your team pulls and analyzes	The system delivers conclusions
Timing	Weekly or monthly, backward-looking	Continuous, forward-looking

The gap between these two columns is where money disappears. Not in dramatic collapses — in slow, invisible leaks that compound week after week.

What decision intelligence actually does

Three things your current tools don't.

It connects your data sources automatically. ERP, accounting, CRM, email, invoicing, banking — decision intelligence pulls from all of them and builds a unified picture. No more waiting for someone to manually cross-reference reports on a Friday afternoon.

It catches anomalies before they become crises. A customer stretching payment terms from 45 to 72 days. A raw material cost trending up 2% per month. A delivery delay pattern from a specific supplier. These signals cost real money when they go unnoticed for three weeks. A decision intelligence system catches them on day one.

It tells you what to do, not just what happened. This is the part that matters most. Instead of "margin on Product Line C dropped 4%," you get: "Margin on Product Line C dropped 4% because resin costs from Supplier X increased. Supplier Y offers equivalent spec at 7% lower cost. Current contract allows renegotiation within 30 days. Recommended action: request quotes from Supplier Y this week."

The difference between knowing something went wrong and knowing exactly what to do about it — that's the entire value proposition.

The math on slow decisions

Here's a concrete scenario. Your purchasing team spots a raw material price increase. It comes up in the weekly ops meeting. The operations director raises it with the CFO. The CFO runs numbers in Excel. A decision gets made two to three weeks later.

During those weeks, every production run using that material is less profitable than your pricing assumes. At 500 units per day with EUR 2 per unit margin erosion, that's EUR 7,000 per week walking out the door. Two weeks of decision latency: EUR 14,000. On one material. On one product line.

Now multiply across every category: pricing adjustments, supplier negotiations, customer retention, production scheduling, compliance deadlines. The compound cost of slow decisions in a EUR 20M manufacturer typically runs 3-7% of annual revenue. That's EUR 600K to EUR 1.4M per year — not from bad decisions, but from late ones.

82% of Italian SMEs have no AI solution at all

That number comes from AICIM's October 2025 survey. It means the vast majority of Italian manufacturers are making every decision the same way they did a decade ago: manually, slowly, and with whatever information someone happened to pull together.

This is both a vulnerability and an opening. The companies moving now aren't early adopters chasing hype. They're the ones who looked at their margin trends and decided they couldn't afford another year of flying blind.

Three forces are making this urgent:

• Regulatory pressure is real and accelerating. CBAM is live. CSDDD transposition deadlines are approaching. CSRD requirements are cascading from large enterprises down to their SME suppliers. Each one demands better data and faster documentation. SMEs that can't produce it face penalties and lost contracts.
• Margins are getting squeezed from both sides. Input costs up, pricing pressure down. The companies that spot margin leakage three weeks earlier will outperform those that don't. Period.
• The technology finally fits SME budgets. Two years ago, this kind of system required a data science team and a seven-figure budget. That barrier is gone.

What this looks like on a Monday morning

You open your inbox. There's a weekly intelligence brief. It tells you:

• Three open invoices at highest risk of late payment, based on each customer's actual payment behavior — with a recommended follow-up action for each
• A supply chain flag: one of your key suppliers appeared in an EU compliance screening last week, with an alternative supplier already vetted
• Product Line C has been running 4% below target margin for three weeks, driven by a raw material cost your pricing hasn't absorbed yet — with a specific price adjustment recommendation
• Your main competitor just filed for a certification in your primary market segment that could affect upcoming tenders

You didn't open a dashboard. You didn't ask your finance team to "pull the numbers." The analysis came to you, already prioritized, already actionable.

Data sovereignty matters more than you think

One thing worth mentioning for European manufacturers: where your business data gets processed is not a minor detail. Customer records, financial data, supplier contracts, competitive intelligence — all of it falls under GDPR, the Data Act, and emerging EU AI regulations.

Platforms built outside the EU may process your data through jurisdictions where foreign government access laws apply. The US CLOUD Act, for example, can compel disclosure of data stored by US-headquartered providers regardless of where the server sits.

Kestevo runs entirely on EU infrastructure. Your data stays in Europe. Full stop.

The gap between knowing and doing

Most manufacturing executives we talk to already suspect they're losing money to slow decisions. They can feel it — the quarterly review where a problem surfaces that someone should have caught two months earlier. The supplier renegotiation that happened a quarter too late.

Decision intelligence closes that gap. Not with more dashboards to check, but with specific, pre-analyzed recommendations delivered to the people who make the calls.

Kestevo does this for manufacturing SMEs with 50-500 employees. We connect to your existing systems, find where money is leaking, and tell you exactly what to do about it — every week, in language a CFO can act on without calling a meeting first.

If that sounds like what your business needs, let's talk.