By Max Schrevelius, Co-founder / Commercial Director
PIM in practice: which product data should you optimise first?
A PIM platform makes it possible to manage, enrich and publish large volumes of product information from one central place. But once PIM is in place, a different question tends to surface: where should your team start improving?
Not every product needs the same level of attention at the same time. And if you try to make the entire catalogue complete in one go, the workload usually grows faster than the team can handle.
A more practical approach is to prioritise product content based on the role products play within your assortment.
That way, PIM does more than help you manage product data. It also helps you focus effort where improvement is likely to create the most value right now.
Read more about what a PIM system is.
Why making everything complete doesn't work.
A common approach is to work through product data systematically.
Supplier A first, then supplier B.
Finish category one before moving to category two.
Or give every product the same amount of content.
It sounds logical, but a product catalogue never stands still.
New products are added. Suppliers update specifications. Prices change. New channels introduce new requirements. And products that were complete last month may need attention again today.
In practice, there is rarely a point where the entire catalogue is simply 'finished'.
So the more useful question is not:
How do we make everything complete?
It is:
Which products deserve our attention first?
Start with your assortment strategy.
Not every product serves the same purpose.
Some products attract traffic. Others contribute more margin. Others increase basket size through cross-sell or upsell.
You can use those commercial roles to decide where product content deserves attention first.
A simple model is to divide products into three groups:
- products that generate traffic;
- products that drive margin;
- products that support additional sales.
The exact balance will vary by business. The point is not to create rigid categories, but to avoid treating every product as equally important.
Read more about Assortment Scaling.
1. Products that generate traffic.
Some products primarily help attract visitors.
These might be well-known brands, popular models, products with high search demand or items that are important within a specific category.
For these products, stronger product content can improve discoverability and help customers reach the right product faster.
That may include:
- clear product titles;
- relevant attributes;
- complete category information;
- unique product descriptions;
- strong imagery;
- consistent search and filter attributes.
If traffic is underperforming in a category, improving these products may create more value than enriching the catalogue at random.
2. Products that drive margin.
High traffic does not automatically make a product commercially important.
Some products matter because they contribute disproportionately to margin. This may include private-label products, own-brand ranges or items bought on particularly favourable terms.
For these products, the content challenge is different.
The goal is not just discoverability. It is about making the product's value clear enough to influence the buying decision.
That may mean:
- sharper USPs;
- clearer comparisons with alternatives;
- stronger articulation of product benefits;
- better imagery;
- relevant product relationships;
- clearer positioning within the category.
If margin is under pressure, prioritising these products may be more effective than improving products that already perform well.
3. Products that support additional sales.
A third group consists of products that create value mainly in combination with other items.
Examples include:
- accessories;
- spare parts;
- consumables;
- add-on services;
- compatible products.
An extra battery for a laptop is only useful if it is clear which models it fits.
For this group, the quality of individual product content is only part of the picture. The relationships between products matter just as much.
Which products belong together?
Which accessories are compatible?
Which add-ons make sense for which main product?
A well-structured data model makes those relationships easier to manage and reuse.
Read more about Data Model & Structure.
The right starting point depends on the problem.
This model becomes most useful when you connect it to a specific commercial or operational challenge.
Not enough traffic?
Start with the products and categories that are meant to attract visitors.
Check whether titles, attributes, category structures and content reflect the way customers search and compare products.
Margin under pressure?
Identify the products that matter most for margin and review whether their value is communicated clearly enough.
Low average order value?
Look at product relationships, accessories and complementary items.
In this case, improving cross-sell and upsell structures may create more value than adding more copy to products that already perform well.
Seasonal peak approaching?
Prioritise categories before demand arrives.
For seasonal ranges, optimising content after the sales window has already started adds limited value.
Too much manual correction?
Then the issue may not be the product content itself.
If teams repeatedly correct, map or complete the same supplier data, it may be more effective to address the source of that work first.
Read more about Supplier Data Onboarding.
PIM makes priorities actionable.
Your assortment strategy determines which products deserve attention. PIM helps turn that priority into an operational workflow.
Products can, for example, be segmented by:
- category;
- brand;
- supplier;
- channel;
- completeness;
- status;
- commercial priority.
From there, workflows, validations and bulk actions can help teams execute the work in a more focused way.
That creates an important shift.
Without prioritisation, PIM can become a large collection of products all competing for attention.
With prioritisation, it becomes a working environment aligned with what the business is actually trying to achieve.
Read more about Workflows & AI.
Completeness is not the same as quality.
PIM teams often track completeness: what percentage of required fields has been filled in?
That is useful, but it is not enough.
A field can be populated and still add very little value.
A product description can exist without being distinctive.
An attribute can be present but use the wrong unit.
An image can be available but be unsuitable for a specific sales channel.
Product quality should therefore be assessed in context.
For a traffic-driving product, discoverability may matter most.
For a margin-driving product, persuasive content may be more important.
For an accessory, an accurate product relationship may be critical.
The definition of 'good product data' depends partly on what the product is expected to do.
Measure whether your prioritisation works.
A prioritisation model is only useful if you can see whether it improves outcomes.
A few metrics can help.
Time to market
How quickly do priority products move from incoming source data to publication?
Completeness
Is the required information available for this product group?
First-time-right
How many products can move through the process without additional correction rounds?
Manual touches
How often does someone need to manually adjust a product before it is ready?
Commercial outcome
After optimisation, do you see changes in traffic, conversion, margin or average order value?
Not every improvement can be attributed directly to product content. But combining operational and commercial metrics makes it easier to see which product groups consistently require attention and where improvements appear to have an effect.
Read more about Reporting & Analytics.
AI mainly changes what you prioritise.
AI makes it possible to automate more standard product-content tasks.
For example:
- generating product descriptions;
- translation;
- attribute extraction;
- classification;
- identifying missing information;
- quality checks.
That does not make prioritisation less important.
It makes it more important.
As routine tasks become faster and cheaper, the question becomes where human attention still adds the most value.
Teams no longer need to enrich every product manually. They can focus on exceptions, commercially important products and cases where judgement still matters.
Read more about AI Product Data Automation.
Good prioritisation sometimes starts before PIM.
Not every problem that becomes visible in PIM should be solved in PIM.
Imagine that products from supplier A repeatedly arrive with incomplete technical specifications.
You can ask your team to fill in the missing information after every import. Or you can investigate whether validation, mapping or enrichment can happen earlier in the product data flow.
That leads to a useful principle:
Solve recurring problems as early in the flow as possible.
Doing so prevents the same correction work from reappearing for every new product.
At ConnectingTheDots, that is an important principle:
Good product data starts before your PIM.
Read more about Import & Onboarding.
Conclusion: don't optimise everything at once.
A PIM platform gives you the ability to manage large volumes of product information centrally.
But it does not automatically decide where your team should spend its time.
For that, you need a prioritisation model.
Look at the role products play in your assortment. Decide which business objective matters most right now. Then configure PIM, workflows and automation so that the relevant product group gets attention first.
That produces a better question than:
"Which products are still incomplete?"
A more useful question is:
"Which improvement, on which products, creates the most value right now?"
Used consistently, that question turns product content management from an endless backlog into a more deliberate part of assortment strategy.
Frequently asked questions about PIM in practice.
A PIM system helps organisations manage, enrich and distribute product information from a central environment.
The goal is not simply to store data in one place, but to make product information easier to control and reuse across teams and channels.
Read more about Product Information Management.
Not necessarily at the same time or according to the same definition.
The information required depends on the product, category and channel. In practice, it is often more effective to combine completeness with commercial priority and channel requirements.
PIM can support data quality through structure, validation rules, workflows and completeness checks.
However, overall data quality also depends on the quality of source data and on how responsibilities are organised across the business.
AI can support repetitive tasks such as classification, translation, enrichment and quality checks.
This allows human attention to shift towards exceptions and products where commercial or editorial judgement matters most.
Yes. A PIM platform can be integrated into a broader ecosystem that includes ERP, DAM, webshops, marketplaces and other systems.
The integrations you need depend on where product data originates and where it needs to be used.
Read more about Integrations.
Which products deserve attention first?
See how your product data moves from source to sales channel, and where better prioritisation, automation or structure could have the greatest impact.
Or book a demo.