If you haven't decided in advance which system gets the final say on where an order came from, you'll never know for certain which campaign generated it: only 21.1% of Italian SMEs use a CRM, against 56.5% of large companies (Istat, "Imprese e ICT", 2025 reference year, press release of 15 December 2025) — and the CRM is exactly the tool that should record that origin.
It always plays out the same way. A company has an active campaign on Google Ads, one on Instagram, a freshly redesigned website, and a newsletter running in parallel. Then an order comes in. Who brought it?
The CRM, if there is one, says "direct" because the site's form never captured where the visitor came from. Google Ads claims the conversion because it saw the click. Instagram claims the same conversion because it showed the ad three days earlier. None of the three is wrong: nobody ever decided how to track the path, so each one only reports the part it saw.
The problem grows with every channel added in parallel: Internet advertising spend in Italy reached €6.2 billion in 2025, with a 2026 estimate of €7 billion, +12% (Osservatori Digital Innovation, Politecnico di Milano, Internet Media Observatory, 2026 release). More channels mean more versions of the same sale to reconcile.
What "tracking from campaign to order" means
Tracking means linking an upstream event (a click on an ad, opening a newsletter) to a downstream event (a contact, an order) through a reference that survives the handover between systems. The most common mechanism is the UTM: a tag added to a campaign's link that tells the site where the visitor came from.
Attribution is the rule that decides, when several channels touch the same customer before the order, which one gets the credit: last-click gives full credit to the last channel seen; multi-touch spreads it across the channels involved. Native tools (Google Ads, Meta Ads) each apply their own rules, only to their own clicks — which is why two platforms can both claim the same sale without either one lying.
The system of record for the origin data is whichever system gets the final word when the versions disagree — usually the CRM, because it's the only one that also sees whether that contact actually became a paying customer.
How many systems tell part of the story?
In an SME running several channels at once, each system only sees the slice of the journey it's involved in, and attributes the sale according to its own logic:
| System | What it sees | What it misses | How it attributes the sale |
|---|---|---|---|
| CRM | The contact, and the "how did you hear about us" field if it exists | Earlier clicks, unless passed as a form parameter | Often "direct" by default, if the form doesn't pass the origin |
| Google Ads | The click on its own ad and, with the tag installed, the on-site conversion | Clicks on other platforms in the same journey | Last-click on its own ads |
| Meta/Instagram Ads | Clicks and, partly, views in the days before the event | Events outside its own attribution window (typically 1-7 days) | Even a view without a click, within its window |
| Website (analytics) | The session and channel, if the link carries a UTM parameter | What happens after the contact leaves the site | Per session: two visits from the same contact are two different journeys |
Note: the table describes typical tool behaviour; attribution settings are configurable and can vary between companies.
Only 41.9% of Italian SMEs analyse their own business data using internal staff or external organisations, against 83.6% of large companies — up from 25.7% and 74.1% respectively in 2023 (Istat, "Imprese e ICT", 2025). The gap has narrowed but stays wide: most SMEs still read each channel on its own, with no point where the versions get compared.
How to set up minimal tracking without a data team
You don't need a data warehouse to move past a default "direct". Starting from level zero, you need three levels, each building on the one before:
| Level | What it takes | What question it answers |
|---|---|---|
| 0 · No tracking | — | None: every contact shows up as "direct" |
| 1 · UTM + form field | A UTM link per campaign, a mandatory field on the form | Which campaign brings in contacts (not yet orders) |
| 2 · CRM as system of record | The CRM records the source at first contact and keeps it until the sale closes | Which campaign brought each closed order |
| 3 · Integrated ads | The CRM sends sales outcomes back to the ad platforms, which correct their attribution | Which channel produces customers who actually buy, not just clicks |
Level 2 solves most of the cases described above: it just takes the CRM being the system everyone else defers to on origin, the same logic used for the customer record in A single customer record: how to build one in an SME. Here it is applied to order origin instead of identity data.
How to put it into practice, step by step
- Choose the system of record for origin. In most SMEs, that's the CRM, since it's the only one that follows the contact through to a closed sale.
- Standardise your UTMs. A fixed convention (source, medium, campaign name) for every published link, including organic posts, not just paid campaigns.
- Add a mandatory field to the contact form that captures the UTM parameter or, when there's no UTM, asks directly "how did you hear about us?"
- Connect the CRM and ad platforms so the original source doesn't get lost at the first human handover (a salesperson calling back, a quote written by hand).
- Review contacts marked "direct" every month. If they're a small share, tracking is working; if they're the majority, part of the chain is still leaking.
If there are more than two systems to connect, the same premise from Integrating CRM and ERP in an SME: where to start applies: decide who has the final word first, then connect the tools, not the other way round.
When it isn't worth it
This is my own judgement, not a figure: with a single active acquisition channel — say, word of mouth plus one Google Ads campaign — building multi-touch attribution is time spent solving a problem that doesn't exist yet. A form field and a monthly glance at the CRM are enough; more sophisticated tracking only pays off from three parallel channels upwards, because that's where the versions really start to contradict each other.
What it costs
There's no one-size-fits-all figure here either: it depends on how many systems are involved and how much is already configured. The spend breaks into three parts:
- an analysis that maps active channels, existing systems, and how many contacts currently show up as "direct" without really being so;
- an implementation that standardises UTMs, adds the form field, and makes the CRM the system of record for origin;
- a monthly review that checks anomalies as the advertising tools in use change.
Anyone working on this kind of project, this author included, typically quotes the analysis alone at a fixed scope before pricing the rest: it's the only way not to promise a price on a problem that hasn't been measured yet.
Frequently asked questions
Do you need a CRM to track where an order came from?
Not necessarily, but without a system that keeps the source from first contact through to the closed sale, every report stops at clicks and never reaches the order. A manually updated spreadsheet can work under a few dozen contacts a month; beyond that, it becomes more work than the CRM it's meant to replace.
What is multi-touch attribution and does an SME need it?
It's the method that spreads credit for a sale across several channels touched in the same journey, instead of giving it all to the last click. It makes sense with three or more active channels; with one or two, the difference from last-click is minimal and doesn't justify the extra complexity.
Why do Google Ads and Meta report different conversion numbers for the same sale?
Because each platform only attributes its own clicks or views, within a time window it sets itself, without seeing what the other platform did in the same journey. Adding the two platforms' numbers together therefore counts the same sale more than once.
Where do you start if the company has never tracked anything?
With the "how did you hear about us?" field on the contact form: it costs an afternoon of work and makes visible, from the first month, how many contacts really have no known source. It's the step that comes before any conversation about a CRM or multi-touch attribution.
In short
An SME that doesn't track the path from campaign to order doesn't have a data-shortage problem: it has three or more systems each telling a true part of the same story, with no rule for deciding which one to trust. The fix isn't buying more tools — it's deciding which system, usually the CRM, gets the final word on origin, and connecting the others to it. Same principle as the customer record and CRM-ERP integration: one system of record, the rest follow.
If your company runs several channels and nobody knows for certain which one brings in the orders, the systems analysis is built exactly to map that situation before proposing solutions. The same system-of-record principle underlies how I set up a CRM: see CRM for SMEs: what it is and when you need one. The first step is a thirty-minute call to walk through which channels and systems you use today.
Sources
- Istat (the Italian national statistics institute), "Imprese e ICT", 2025 reference year, press release of 15 December 2025 (enterprises with at least 10 employees; CRM: 21.1% of SMEs and 56.5% of large enterprises; data analysis with internal or external staff: 41.9% of SMEs and 83.6% of large enterprises, up from 25.7% and 74.1% in 2023; social media: 59.0% of enterprises).
- Osservatori Digital Innovation, Politecnico di Milano, Internet Media Observatory, 2026 release (Internet advertising in Italy: €6.2 billion in 2025, €7 billion estimated for 2026, +12%, 53% of total advertising spend).
- For European comparison: Eurostat, indicator isoc_eb_iip (2025), CRM software use among EU enterprises, 28.51% overall and 24.69% among small enterprises (10-49 employees). This is a different universe from Istat's SME bracket (10-249 employees), so the comparison is indicative, not point-for-point.
There is no official national statistic on the share of Italian SMEs that correctly attribute sales to their originating campaigns: where no verified data exists, this article says so explicitly and distinguishes between Istat data, Eurostat data and the author's own judgement.
Sergio Selvelli, 22 September 2026.