Case studies / Advisory
Advisory · CPG ecommerce architectureWe didn't run the account. We taught the team who still does.
American Licorice Company's marketing team was taking Sour Punch direct to consumer and wanted the Merchant Center and Google Ads architecture right before they built it. Two calls a week for three months. They've run it themselves ever since.
A household candy brand, going direct to consumer for the first time.
American Licorice Company has been making confectionery for over a century, and Sour Punch is one of its best known brands, sold nationwide through retail.
We were brought in as the in-house marketing team built a direct to consumer channel from scratch, advising on the Merchant Center and Google Ads architecture before they built it themselves.
Some accounts shouldn't be outsourced. They should be built properly and handed over.
A national CPG brand already has a marketing team, a merchandising calendar, and a retail strategy that a D2C store has to respect. Handing paid media to an outside agency means all of that context has to be explained, repeatedly, forever. The alternative is to get the architecture right once, with the people who hold the context, and let them run it.
Fast to launch, permanently dependent.
Someone else builds it and keeps it. The knowledge sits outside the business, the retail and promotional context has to be transferred every cycle, and if the relationship ends the team inherits an account nobody internally understands.
- Who holds the knowledge The agency
- Brand context Explained repeatedly
- Monthly cost Ongoing
- If it ends Capability leaves
Slower to launch, permanently theirs.
The internal team builds the account with guidance, so they understand every decision in it. The context never has to be transferred because it never leaves. The cost is a fixed advisory period rather than an indefinite retainer.
- Who holds the knowledge The team
- Brand context Already there
- Cost Three months
- Afterwards They keep running it
A household brand, starting D2C from nothing.
American Licorice Company has been making confectionery for over a century, and Sour Punch is one of its best known brands. The product was everywhere in retail. Selling it directly online was a new channel, and the marketing team wanted to build it rather than outsource it.
Four things made that worth doing carefully.
Merchant Center structure is hard to undo
Account setup, product data standards, and feed architecture shape everything built on top of them. Getting them wrong is not a bad month, it's a rebuild, and often a suspension risk in the meantime.
A known brand has different search behaviour
People already know Sour Punch. That changes the balance between brand and non-brand search, and it means brand terms need defending rather than discovering.
The team held context we didn't
Retail relationships, promotional calendars, packaging formats, and pricing constraints all sit inside the business. Any architecture built without that knowledge would have needed rebuilding once it met reality.
They wanted to own it, not rent it
The goal was explicitly capability, not coverage. Anything we built for them rather than with them would have failed that test.
Four campaign types, and why each existed.
Every structure came with the reasoning, so the team could adapt it later without asking us.
Merchant Center, built to be built on
Account structure, product data standards, and feed architecture designed first, including how products should be described, categorised, and grouped so the catalogue could grow without the structure needing to change underneath it.
A foundation the team could extend themselves.
National Shopping as the primary channel
Shopping campaigns structured to carry the catalogue across the country, with product grouping designed so performance could be read by product type rather than as a single blended number.
The main revenue channel, measurable by product.
Brand and non-brand Search, kept apart
A brand campaign to defend searches for the name the brand had spent decades building, and separate Search campaigns on specific product and category terms. Blended together, brand traffic flatters the whole account and hides whether the non-brand work is doing anything.
Brand protected, non-brand honestly measured.
Retargeting split by what the visitor had already done
Two distinct audiences rather than one: previous customers, reached with new promotions, and people who added to cart and never checked out. A returning buyer needs a reason to come back; an abandoner needs the objection that stopped them removed. The same ad serves neither well.
Two audiences, two messages, separately measured.
Hands on the keyboard where it helped
Alongside the weekly sessions, direct support with implementation and integration, the technical steps that are faster to do together than to describe. Advisory that stops at the slide deck leaves the hardest part undone.
A working account, built by the people who'd keep it.
They still run it. That's the result.
We're not publishing traffic, sales, or return figures for this engagement, because we didn't run the account and don't hold them.
What we can point to is more durable than a number. Three months of advisory, and the internal team has been running the account themselves ever since, building on the same architecture, without us.
That's the actual test of advisory work. A system that only functions while the consultant is attached hasn't been transferred, it's been rented. One that a client's own team extends for years afterwards was built correctly and explained properly.
It's also the honest argument against hiring us for everything. For a business with a capable marketing team and context an outsider would take a year to absorb, the right engagement is often a short one.
Four things for any in-house team going D2C.
Get the foundation right before the campaigns
Merchant Center structure and product data standards are the hardest things to change later. Decide them deliberately, before anything is built on top.
Separate brand from non-brand
If your brand is already known, brand search will make the whole account look excellent. Split it out or you'll never know whether the rest is working.
Retarget by behaviour, not as one list
A past customer and a cart abandoner need opposite messages. Treating them as one audience wastes the cheapest conversions available to you.
Buy expertise, not dependency
If your team holds context an agency would take months to learn, a fixed advisory period usually beats an indefinite retainer.
What went into it.
Five of our services, delivered as advisory.
Do you need an agency, or a few months of help?
If you have a marketing team and they'd rather own the channel than outsource it, the right engagement might be a short advisory one. Tell us what you're building and we'll say honestly which of the two you need, including when the answer is neither.
No obligation · We'll tell you if you don't need us