Deep vs Surface Customer Enrichment: What Is Actually in an Ecommerce Customer Profile

Two ecommerce enrichment tools can return a profile for the same customer and mean completely different things by the word "profile."

One returns a row: name, email, a persona label like "fitness enthusiast," maybe an income band and a follower count. The other returns a researched dossier: which platforms the customer is active on, what they actually post about, which brands and communities show up in their content, their household and demographic context, and a link to every source behind each claim. Both are called "customer enrichment." They are not the same product, and the gap explains most of the confusion when brands compare tools and pricing.

I'm the co-founder of Mercana, so read this with that context. It's also not armchair theory for us. Before Mercana, I (Daniel Lee) worked on the enrichment algorithm and infrastructure at Clay, specifically the enrichment that powers ecommerce actions, so the matching, verification, and source-tracking problems in this post are ones I've been elbow-deep in for years. My co-founder, Vijay Sridhar, came out of product engineering at Meta. We built Mercana's enrichment pipeline ourselves, from the ground up, so when I draw the line between a lookup and a research job, it's from having built both sides of it.

This is a guide to the distinction that matters most when you evaluate any customer intelligence platform for a DTC or consumer brand: surface enrichment versus deep enrichment, and how to tell which one you are actually buying.

The short version

Surface enrichment is a database lookup. It matches your customer to a record and appends a fixed set of pre-computed fields. It is fast and cheap per customer, and it is genuinely useful when all you need is a category tag or a quick "is this someone notable?" flag.

Deep enrichment is a research job. For each customer, it searches public social and web sources, reads the content the person has published, verifies the identity match, and assembles a profile with far more data points, each tied back to a source. It costs more per customer because more happens per customer.

The practical test: after enrichment finishes, can you make a decision, or do you have another spreadsheet to interpret?

What surface enrichment gives you

Surface enrichment is built around a static database. The tool takes an email or name, finds the best matching record, and returns whatever fields that record already holds. For ecommerce that usually looks like:

• Name, email, and location
• A persona or segment label (for example "gardener," "fitness enthusiast," "new parent")
• A follower count on one or two platforms
• A demographic estimate or two (age band, income band)
• A notable-buyer flag if the person is a public figure

This is fine for what it is. If your only question is "which customers are celebrities or big accounts," a surface tool answers it quickly and inexpensively. The limitation is that a fixed field is the end of the road. The label "fitness enthusiast" does not tell you what this customer trains for, which brands they wear, whether they post content you could amplify, or whether they belong in a creator shortlist versus a lifecycle branch. Someone on your team still has to go find that out.

What deep enrichment gives you

Deep enrichment treats every customer as a small research task instead of a database key. Mercana enriches each customer with 200+ data points and, critically, analyzes the customer's public social context rather than just their profile stub. Concretely, that spans four layers:

1. Identity across platforms. LinkedIn, Instagram, X, TikTok, and Facebook profiles, with follower counts, bios, and engagement metrics, so you see reach on every channel rather than one.

2. Public content, not just the handle. This is the layer surface tools skip. Mercana analyzes public social profile context, including tags, posts, comments, and images where available. That means the profile can reflect what a customer actually posts about and which brands or themes recur in their content, not only that an account exists.

3. Household and demographic context. Home value, location, occupation, and interests, so a customer is placed in a real-world context, not just an online one.

4. Commerce signals from your own store. Order value, subscription status, and lifetime value, joined to everything above so identity and behavior sit in one place.

Every one of those data points is tied back to the customer record with confidence scores and source links. You are not asked to trust a black box; you can see why a match was made and audit it before you act.

Side-by-side

	Surface enrichment	Deep enrichment (Mercana)
Method	Database lookup, fixed fields appended	Per-customer research across public social and web
Social data	Handle plus follower count	Profiles across 5 platforms, plus posts, tags, comments, and images where available
Real-world context	Rough demographic estimate	Home value, location, occupation, interests
Personas	Pre-set category label	Personas built from your actual customers, with lookalike modeling
Evidence	Field value only	Confidence score and source link on each enrichment
Data points	A handful of fixed fields	200+ per customer
After enrichment	You interpret and export	You get analysis, routed audiences, and recommended actions

Why depth changes what you can ask

The reason depth matters is not "more fields are nice." It is that a fixed set of fields can only answer the questions someone anticipated in advance. Deep, content-level enrichment lets a team ask open-ended questions of its own customer base, the kind that actually drive creator, lifecycle, and acquisition decisions:

• What are my customers posting about, and which outside brands show up most often in their content?
• Which schools, occupations, or communities over-index among my buyers?
• Who in my customer file has real reach on Instagram or TikTok, and what do they post?
• How do my highest-LTV customers differ from my lowest-LTV customers by identity, not just by spend?

None of these are profile fields. They are analysis jobs: gather evidence across thousands of customers, aggregate it, and explain what it means. A surface tool cannot answer them without your team doing the export, the pivot, and the interpretation. This is the same reason deep enrichment feeds influencer and creator discovery that a follower-count field cannot, and why it underpins reliable VIP detection instead of guessing from spend alone.

Depth is not the enemy of accuracy

A fair worry about deeper enrichment is that pulling from more sources means more chances to match the wrong person. The opposite is true when every match is verified instead of assumed. Because Mercana attaches a confidence score and source links to each enrichment, a low-confidence match is visible rather than silently wrong, and your team can audit any profile before acting on it. Across more than 5M profiles enriched, matching runs at 90%+ accuracy with that evidence attached. Source transparency is what makes depth trustworthy; a field with no provenance is the harder thing to defend.

What this means for buying

When you compare customer intelligence tools, "does it enrich customers?" is the wrong question, because the answer is always yes. The useful questions are about depth and what happens after:

1. How many usable data points come back per customer, and do they include public content signals and household context or only a category tag?
2. Is each field sourced? Can you see why a match was made?
3. How much manual work remains before the enrichment becomes a decision or a routed audience?

A cheaper per-customer rate often buys a shallower profile, which moves the real cost downstream to your team's time. A higher rate can buy a researched profile that arrives ready to act on. That trade-off, not the sticker price alone, is what to weigh. I break the pricing side of this down in what customer enrichment actually costs, and the head-to-head product comparison lives in OuterSignal vs Mercana.

Deep enrichment is the input; turning it into routed audiences, lifecycle segments, creator outreach, and measured workflows is the point. If you want to see the depth on your own customers, Mercana offers a free 1,000-customer audit: connect your store and get back a validated segment, the evidence behind it, and a routed action, so you can judge the difference on real data instead of a feature list.

Frequently asked questions

What is the difference between surface and deep customer enrichment? Surface enrichment appends a fixed set of fields from a database lookup. Deep enrichment runs a per-customer research job across public social and web sources, reads what the customer actually posts, verifies the match, and returns 200+ data points with confidence scores and source links. Surface enrichment tells you a category; deep enrichment shows you the person and the evidence.

What data does deep customer enrichment collect? Identity across LinkedIn, Instagram, X, TikTok, and Facebook (follower counts, bios, engagement), public social content (posts, tags, comments, and images where available), household and demographic context (home value, location, occupation), and commerce signals (order value, subscription, lifetime value), all tied back with confidence scores and source links.

Why does deeper enrichment matter for a DTC brand? Because revenue decisions depend on what a specific customer cares about and how influential they are, not just which category they fall into. Deep enrichment answers open-ended questions a fixed field cannot, which is the input to creator outreach, lifecycle segments, and paid-media audiences.

Does deep enrichment mean scraping private data? No. It works from public social and web context tied to customers you already have a commercial relationship with, plus your own store's order data. It reads publicly available posts, bios, and profile signals, not private messages or gated accounts.

How do I evaluate enrichment depth before buying? Run the same customer list through each tool and compare outputs: usable data points per customer, whether public content and household context are included or only a persona tag, whether each field carries a source link, and how much manual work remains before you can act.