Home Brand & Non-Branded Product Matching

Improve AI matching accuracy for home brand, non-branded, and lesser-known brand products by tagging them with a brand classification. This tells the matching algorithm to focus on what the product does rather than what brand it is.

Why Tag Products?

Nesika AI uses advanced algorithms to match your products with competitor equivalents. By default, brand name is a key matching criterion — this works well for established brands like Ajax, Dettol, or Cold Power.

However, for home brand, non-branded, or lesser-known brand products, this can cause problems:

Poor matches: “Woolworths Essentials Floor Cleaner 750ml” may not match “Coles Lemon Floor Cleaner 750ml” because the brands differ — even though they are functionally the same product.
No matches found: Generic or unbranded products may get no matches at all because there is no brand to match on.
Low confidence: A lesser-known brand like “Earth Choice Dishwashing Liquid 500ml” may get a lower confidence score than it deserves when matched against more recognisable alternatives.

Quick Fix

Adding a brand classification tag tells the algorithm to de-emphasise brand and focus on functional equivalence — resulting in better matches for these product types.

Brand Classification Tags

There are three recognised tag values. Add the one that best describes your product:

home-brand

Retailer Private Label

Products sold under a retailer's own brand name. You want to find the closest equivalent product at other retailers, even if those retailers sell it under a different house brand.

Examples: Woolworths Essentials, Coles Homebrand, Aldi Diplomat, IGA Black & Gold, Costco Kirkland Signature

non-branded

No Brand Affiliation

Truly generic or unbranded products with no brand on the packaging. Matching is based purely on product type, size, weight, and description.

Examples: Generic rubber bands, unbranded cleaning sponges, plain envelopes, bulk hardware supplies

2nd-tier-brand

Lesser-Known / Budget Brand

Products from a real but lesser-known brand. The AI may struggle to match these because competitors don't carry the same brand, even though they carry functionally identical products.

Examples: Earth Choice, Sard Wonder, Hercules, Aim, or any budget-tier brand that competitors may carry under different names

Tag Format

Tags are case-insensitive and accept spaces or hyphens. For example, “Home Brand”, “home-brand”, and “HOME-BRAND” all work the same way.

How to Tag Your Products

Using the Tags Field (Recommended)

Navigate to the product

Go to the Products page and find the product you want to tag.

Open product editing

Click on the product to open its details, then click the edit button.

Add the tag

In the Tags field, type one of the three classification tags:home-brand,non-branded, or2nd-tier-brand

Save and re-scan

Save the product. The tag takes effect on the next scan — run a new scan to see improved matches.

Using Custom Fields (API / CSV)

If you manage products via the API or CSV import, you can set the brand classification through the brand_classification custom field:

{
  "customFields": {
    "brand_classification": "home-brand"
  }
}

Re-scan Required

Brand classification tags are read at scan time. After adding or changing a tag, you need to run a new scan for the updated matching behaviour to take effect.

What Changes in Matching

When a product has a brand classification tag, the AI matching algorithm adjusts its behaviour in several ways:

✓ Brand Mismatches Not Penalised

Normally, if your product brand doesn't match a competitor's brand, the confidence score drops. With a brand classification tag, this penalty is removed.

✓ Focus on Functional Equivalence

The AI shifts focus to product type, category, key attributes (e.g. “antibacterial”, “lemon scented”), weight/volume, quantity, and variant characteristics.

✓ Perfect Match Possible Across Brands

A cross-brand match can receive a “Perfect Match” classification if the products share the same core function, equivalent characteristics, size within ±15%, and the same quantity.

ℹ Core Purpose Test Still Applies

The algorithm never matches products that serve fundamentally different purposes. A home-brand floor cleaner will not match a dishwashing liquid, even with brand de-emphasised.

Standard Products Unchanged

Products without a brand classification tag continue to use the standard matching algorithm with full brand weighting. No existing behaviour changes for your other products.

Verifying Improved Results

After tagging a product and running a new scan, look for these improvements:

Competitor matches appearing from different brands that previously were not matched
Higher confidence scores for cross-brand matches
Matches classified as “Perfect Match” or “Similar Product” across different brands
Fewer “Not Found” results for generic or home-brand products

Before & After Comparison

Note down your current match results before adding the tag, then compare after the next scan completes. This makes it easy to see the improvement.

Frequently Asked Questions

Do I need to tag every product?

No. Only tag products where you are seeing poor matching results due to brand mismatches. Standard branded products (Coca-Cola, Dettol, etc.) should be left untagged — the default matching works best for them.

Which tag should I use if I'm not sure?

If the product is your retailer's own brand, use home-brand. If it has no brand at all, use non-branded. If unsure, start with 2nd-tier-brand — it provides the lightest brand de-emphasis.

Can I remove the tag later?

Yes. Removing the tag restores standard matching behaviour. The next scan will use the default brand-sensitive matching for that product.

Does tagging affect my existing matches?

Not immediately. Tags only take effect on the next scan. Your existing matches and price history are preserved.

Will this increase my scan costs?

No. Brand classification is metadata only — it changes how the AI interprets results, but does not add extra processing steps or API calls.

Can I tag products in bulk?

Yes. Use the CSV import with a brand_classification custom field, or update tags via the API for multiple products at once.

← Previous: Products Guide Next: Competitors Guide →