SDS Authoring and Chemical Compliance Software: Custom Build Versus Off the Shelf

What the off the shelf products actually do well

Sphera and Verisk 3E carry regulatory content that would take a decade to assemble: substance data, jurisdiction rules, the standardised hazard and precautionary statements with their official translations, and the people who read every revision as it is published. Chemwatch is strong on substance data and widely used. The SAP Environment, Health and Safety module has one structural advantage nothing else has, which is that it already lives inside the system holding your recipes. Cority and Lisam ExESS both cover authoring competently at mid market scale.

Nobody should build a substance database. That is the clearest line in this category, and any developer proposing one is offering to spend your budget recreating something that already exists and is maintained by people who do nothing else. License the content.

Most manufacturers should also buy the software around it. If you make a couple of hundred formulations, ship into two or three jurisdictions, and your recipe set is stable, an authoring service will produce compliant documents for less than a build costs and will absorb the rule changes on your behalf. Outsourcing authoring entirely is a legitimate answer at that scale, and we say so to companies who came for a quote. It is also worth stating that software does not classify anything. Competent people do, and if you have no regulatory affairs capability in house, a build is the wrong purchase.

Where they stop: the recipe changed and nobody told the document

A production chemist substitutes a surfactant because the original supplier is on allocation. The change goes through as a bill of materials revision, technical approves it, and it runs on Monday. The safety data sheet was authored fourteen months ago from the previous formulation and sits as a file on a shared drive and in a distributor portal. Nobody told regulatory affairs, because a bill of materials revision is a routine event that happens dozens of times a month.

Six weeks later a customer's safety team notices the sheet does not carry a hazard class the new component brings, or a shipment is held because the transport paperwork does not match the declared composition. This is a plumbing failure rather than a knowledge failure, and it is where packaged products stop: the connection to your recipes is usually a periodic extract, and a weekly file of formulations is not the same as knowing formulation 4471 changed on Tuesday afternoon and its documents are now stale.

The second gap is that classification gets remembered rather than recomputed. Under the globally harmonised system, classification of a mixture is calculated from its components, their concentrations and their own classifications using defined rules and cut off values. Every input can change underneath a stored decision. A supplier revises their own sheet and adds a hazard. A jurisdiction adopts a newer revision, as the hazard communication standard at 29 CFR 1910.1200 did when it was aligned to a later revision of the system. A harmonised classification changes under the CLP Regulation. Without a cascade, your finished product classification quietly depends on supplier data that was accurate whenever somebody last read it.

The arithmetic: per document authoring versus a build

This category has a genuine per transaction price, which makes the comparison unusually clean. Authoring services are quoted per safety data sheet, per jurisdiction, per language, and revisions are quoted the same way. Software is priced per product or per authoring seat. So count your documents rather than your products: formulations multiplied by jurisdictions multiplied by languages, plus the revision rate, which is the number people forget.

At 150 formulations into three jurisdictions in two languages, with a handful of revisions a year, you are in the low hundreds of documents and every authoring quote will beat a build across five years. At 900 formulations into fourteen jurisdictions in eleven languages, with a live recipe set, you are producing and revising documents in the tens of thousands, and per document pricing has stopped resembling anything rational.

The crossover lands at roughly 3,000 to 5,000 live document versions, which for most manufacturers means somewhere above 300 formulations once you pass five or six jurisdictions. Below it, buy or outsource. Above it, build the workflow layer around licensed content and keep paying the content subscription, because that is the half that stays maintained without you.

What a custom build actually costs

A focused first release covering an event driven recipe feed from your enterprise system, a classification engine for your home jurisdiction, a managed phrase library and document generation in one or two languages runs $80,000 to $170,000 and ships in 12 to 18 weeks in Digital Heroes delivery experience. A full platform adding further jurisdictions and languages, label generation, transport classification, poison centre notification output and a distribution register runs $220,000 to $500,000 phased over 9 to 18 months.

Data migration runs 10 to 25 percent of build cost and sits high whenever formulations are recorded as free text rather than structured components with concentrations, because that is a data project before it is a software project and it should be priced honestly rather than discovered in week six. Year two onwards runs 15 to 20 percent of build cost annually, covering jurisdiction rule updates, new languages, translation review cycles and label template changes.

What moves the number: the count of jurisdictions, since each has its own required content and its own adoption state. The count of languages, because translation governance is an operating cost as well as a build effort. And label printing integration, which is fiddly and legally sensitive, because the label is the artefact a worker actually reads.

What holds the number down is starting narrow. Take your home jurisdiction, your top product families by revenue, and licensed substance content rather than your own database. Extending to further jurisdictions afterwards is incremental once the classification model and the phrase governance are right, and expensive if they are not.

The four situations where building wins

• Regulatory fit. Supply classification under CLP, workplace communication under 29 CFR 1910.1200, transport classification under the road, air and sea rules, and poison centre notification under Annex VIII of CLP with its unique formula identifier are four determinations that can legitimately differ for one product. Generating all of them from one versioned classification object is the only way they stay consistent.
• Scale economics. Per document authoring across a large multi jurisdiction range, revised whenever a supplier moves, is a bill indexed to exactly the thing your business does. Companies with an active reformulation programme pay it hardest.
• A workflow that is your competitive advantage. Which technical authority approves a hazard determination, how a customer specific variant is handled, what happens when a raw material supplier revises their own sheet. Those rules are yours, and in packaged products they end up as email around the product rather than inside it.
• Integration sprawl across three or more systems. The enterprise system holding recipes, a laboratory information system, a licensed substance database, your label printers and the distributor portals that hold your issued documents. When drift between those is the risk, the connection is the control.

Two of those four together is the threshold we use. One alone usually argues for a better extract schedule or a better authoring contract, and both of those can be arranged in a fortnight rather than a financial year.

How to decide in a week

Test the plumbing rather than the product. On Monday pull the last twenty bill of materials revisions from your enterprise system, with dates. On Tuesday check, for each one, whether the safety data sheet and the label were reissued, and how many days passed. On Wednesday take one product sold into five countries and try to list every current document version and every superseded one, with the date each was issued. On Thursday pick one raw material whose supplier sheet was revised this year and find every finished product containing it.

By Friday you have the only three numbers that matter here: how many recipe changes moved a document, how long the lag was, and whether Thursday's cascade was answerable at all. If nineteen of twenty reissued within days and Thursday took ten minutes, your process is working and you should renew. If Wednesday produced a folder and Thursday produced a shrug, the link between recipe and document is not a control, it is a person hearing about a change.

Then commission the specification first. Digital Heroes runs a paid discovery phase ending in a signed product requirements document covering the change event design, classification versioning, phrase library governance, the document matrix and acceptance criteria, at a fixed price, and it is yours whoever builds it. We are wrong for a manufacturer with no in house regulatory affairs capability, because software cannot make classification decisions, and wrong for anyone expecting our people on your site, since we hold no local office anywhere. We do have more than fifty specialists, over 2,000 delivered projects, our own products including ShopScore, HeroCheckout and Section Vault, and India LLP, US LLC and UK LTD entities so intellectual property assigns under your own law. You meet the named team before signing and can verify us on Clutch, Trustpilot, Fiverr Vetted Pro and D-U-N-S.

Book a 30-minute call with Digital Heroes and get a written plan and a fixed quote within 48 hours.