Quick Answer
Multi-site projects rarely fail because the wrong product was selected. They fail because the same decision was made independently at each site, and the sites did not agree.
The symptoms are familiar: four sites, three different adhesive models, two different coverage figures, and one site that substituted a product because the delivery was late. Nobody made a bad decision; the process allowed good decisions to diverge.
Three controls prevent that:
- A locked specification — one approved model per application, with the substitution rule written down.
- A shared document set — same TDS revisions, same approvals, same language, everywhere.
- A common record — so a failure at site 3 can be compared with site 1 instead of argued about.
1. Where Sites Actually Diverge
| Variable | Why it drifts | Cost of drift |
|---|---|---|
| Substrate | Local materials and local trades | Wrong class specified for one site |
| Climate | Temperature and humidity at application | Cure and open-time problems on one site only |
| Contractor habit | “We always use this one” | Unapproved product, no documentation |
| Local specification | Consultant preference per region | Additional approvals and testing per site |
| Availability | Stock-out at the distributor | Silent substitution |
| Language | Documents in different revisions | Different understanding of the same detail |
Only the last two are supply problems. The rest are governance problems, and they are cheaper to fix before mobilisation than after handover.
2. Lock the Specification, Including the Escape Route
A locked specification has three parts:
One approved model per application. Not a range, not “an MS polymer of equivalent performance” — a code from the current product range. A model code, so that a site cannot interpret its way into a different product.
A stated substitution rule. Something like: no substitution without written approval, and any approved substitute must be accompanied by its own TDS, batch record and, where required, its own test evidence. Without this rule, the default site behaviour during a stock-out is to use whatever is available.
A defined equivalence. If two models are genuinely interchangeable for a given application — for example a standard and a low-modulus grade in the same displacement class — say so explicitly, with the conditions under which the substitution is allowed.
The specification discipline that prevents most of this is the same one used for procurement generally: the building envelope sealant specification checklist shows how substrate pairs and movement classes are written so that they cannot be read two ways.
3. Documents Across Jurisdictions
A multi-site project usually crosses at least one regulatory boundary. Two rules keep the document set manageable:
- One revision register, listing each model with the TDS revision, SDS revision and certificate numbers, in one place. Each site references the register, not a folder of PDFs.
- Per-market sets are separate, and that is normal. Outdoor construction products are assessed against CE, ISO 9001 with ISO 11600, and RoHS in this range; indoor products carry a different set. A site asking for an outdoor product’s indoor marks is a scope question, not a missing document.
Where a market requires a national approval, plan the lead time into the programme. Approvals are applied for per market; they are not transferred with the product, and no supplier should claim to hold one they do not.
4. Supply Risk Is a Selection Criterion
On a single site, a two-week delay is an inconvenience. Across six sites, it is a programme event, and it is the moment unapproved substitution happens.
Three practical controls:
- Batch continuity. Ask for the batch range allocated to the project, and whether it can be held. A project sealed with three production batches is harder to diagnose if something goes wrong.
- Replenishment lead time, measured from the event that actually starts it — deposit receipt, not order acknowledgement.
- A named second source for the highest-volume models, with the substitution rule already applied. Better to approve the alternative in advance than to negotiate it during a stock-out.
The same questions a distributor asks about continuity are relevant here; how to evaluate a sealant supplier for distribution frames them from the reseller side.
5. Training and Supervision
A locked specification is only as good as the people applying it. Three items pay for themselves:
- A one-page application sheet per product: notch or bead size, open time, temperature window, tooling, cure protection.
- A supervisor briefing at each site, with the sheet and the model codes, not a general toolbox talk.
- A first-area sign-off — the first 20 m² or the first 50 m of joint is inspected and signed before production continues.
Where a site’s factory or supplier has to be audited before mobilisation, how to qualify construction adhesive manufacturers in China sets out the checks.
The last item catches more defects than any amount of after-the-fact inspection, because it corrects the crew while they are still able to change their habit.
6. The Shared Record
A single register, identical columns at every site:
| Field | Why it must match across sites |
|---|---|
| Model and batch | Allows cross-site comparison of any failure |
| Area and date | Locates the work in time and space |
| Substrate and preparation | Distinguishes substrate issues from product issues |
| Conditions | Temperature and humidity explain cure anomalies |
| Coverage achieved | Reveals under-application systematically |
| Adhesion test result and failure mode | The single most useful diagnostic field |
When all sites use the same columns, a problem at one site can be tested against the others in an hour. When each site invents its own form, the same investigation takes a week and usually ends in a commercial argument.
7. Governance at a Glance
| Decision | Owner | Evidence required |
|---|---|---|
| Model per application | Central specification team | Approved submittal with model codes |
| Substitution | Central team, in writing | TDS, batch, and test evidence for the substitute |
| Site conditions compliance | Site supervisor | Daily record, temperature and humidity |
| First-area acceptance | Site supervisor and client | Signed inspection sheet |
| Document revision control | Document controller | Register with revision numbers |
| Supply continuity | Procurement | Batch allocation and lead-time confirmation |
Six rows. A project that can fill them in has a controlled specification; one that cannot is relying on luck and goodwill across sites.
Discuss Your Multi-Site Project with INGADA
If you are standardising adhesives or sealants across several sites, send the application list, the substrate types, the markets involved and the programme. We will propose a model per application with the substitution conditions stated, provide the document set per market, and confirm batch allocation and lead times for the volume.
Product selection must be confirmed against each site’s substrate and conditions. Substitution without written approval invalidates the specification.
