Marketing plan

The client

A Canadian brand selling a natural skin conditioning wash for people and animals, sold through its own Shopify store and advertised on Meta.

Business

STANK Products · Quebec

Website

stankproducts.com · Shopify

Range

7 products · the human wash in 1, 2 and 3 bottles, dog, horse, sphynx cat and a family pack

Pricing

$34.50 to $84.50 · CAD

Languages

English and French

Advertising

Meta, $100 a day · no Google advertising yet

Most of the work since 29 September has been behind the scenes: the parts of the site that Google, Meta and screen readers read, rather than the parts customers see. The two changes customers can see, the Shop buttons on the articles and the Skin Guides menu, went in with Julie's OK.

• Done
  Search titles and descriptions for all 7 products, the 3 collections and 8 pages, keeping Julie's own wording where it was already right. These are what Google shows in its results.
• Done
  Image descriptions on every product photo and on 26 photos in the page sections.
• Done
  Product types and tags, and Google categories for the animal products, so Meta and Google can tell the human wash from the animal washes.
• Done
  Structured data for Google: the business, the website, page breadcrumbs and the two articles. The homepage no longer describes itself to Google as a single product.
• Done
  Shop buttons on both articles the ads send people to.
• Done
  The menu: STANK Insights is now Skin Guides, in bold.
• Done
  Meta: every running ad now shows only the human products underneath it. Checked on 2 and 3 October, only the 3 human products appeared.

The call on 2 October set the priorities. Julie's figures on the call: since the launch on 24 September, 23 people had added to cart, 18 reached checkout and about 10 bought. Most buyers choose the two or three bottle bundle.

With margins around 30% before advertising, discounting is not the way to win back the people who nearly bought. The plan works on confidence instead: reviews where the decision is made, shipping and returns stated plainly, nothing slowing the page down, and advertising that brings the people who nearly bought back with more information.

Nothing customers can see changes until it has been approved. Changes to the site's design are tried on a copy of the site first, and every change is logged so it can be put back.

This is the largest gap, and the call put it first. The Reviews page carries a dozen written testimonials, and they are strong, but most are from pet owners and only two or three are about human skin, which is the product the advertising sells. On the product pages the review section sits near the foot of the page, below the before and after photos, and reads "Be the first to write a review", because none have been collected through the review app yet. Its Write a review button is the only way to leave one, and nothing asks buyers to use it.

Where the reviews will show

On the product page, a star rating under the product name and the reviews higher up the page, near the price.

In the cart, short review highlights under the total, plus a scrolling line of them above the cart. The review app's cart widget is already switched on in the theme, so the groundwork is there.

On the Reviews page, a human half and a pet half.

• First
  Ask every buyer automatically. Judge.me is already installed. Switching on its review requests means each buyer is emailed a set number of days after their order ships and can reply with stars, a few words and a photo. This takes over from the personal follow-up emails, and the people who have bought since 24 September can be asked too.
• First
  Use the existing testimonials well: human ones on the human wash, pet ones on the pet washes. Several came from testers who were sent the product, so they show as tester feedback rather than star-rated reviews; star ratings come from buyers through the review app.
• Then
  Show them where the decision is made: the rating under the product name, the reviews up the page, highlights in the cart.
• Then
  Recent purchase notices ("someone just bought the Double Bundle"), built from real orders only. Options will come with their cost; nothing that adds a monthly fee goes in without a yes.

The second thing a buyer needs before paying is to know what happens after. Each line below is what the site shows today, with the change beside it.

Every page is covered by a plain white screen for about five seconds. The page underneath is ready in under a second: measured on 30 September at 0.6 seconds on a computer and 0.8 on a phone, with the white screen lifting at 5.3 to 5.6 seconds. The screen runs on a fixed timer rather than waiting for the page, so every visitor waits the full five seconds, including everyone arriving from an ad.

The header uses the whole bottle label, tagline and small print included, inside a white box on the cream header, at 120 pixels wide on a computer and 90 on a phone, so the small print cannot be read.

The founder-style videos are what sells. From 25 September to 1 October they produced 7 of the 9 purchases Meta recorded: 5 from the Founder Reel and 2 from Skin Won't Heal.

Smaller items that together make the site feel finished.

• Fix
  The French footer still shows [email protected] as its text. The link behind it already goes to [email protected].
• Fix
  The four social icons in the footer link nowhere. Facebook and Instagram get linked, and the LinkedIn and X icons come off unless those accounts are in use.
• Fix
  How It Works has placeholder text (Lorem Ipsum) in its questions section.
• Fix
  The Description box on every product page, the three human ones included, holds the dog wash text: "Designed for dogs with sensitive skin...". Each gets its own (sections 12 and 13).
• Fix
  A homepage sentence ends mid-word: "neutralizes the bacteria making the smel".
• Fix
  Some page addresses do not match their product. The dog wash lives at /products/horse-wash-refill and the family pack at /products/dog-wash. They move to new addresses, with redirects from the old ones so existing links and ads keep working.
• Fix
  The 3 bottle bundle is filed under Acne Treatments in Google's product categories, while the other two human products are under Body Wash. It moves to Body Wash.
• Fix
  The Family Pack lists 64 options for 20 real combinations, because the same three bottles in a different order count as a different choice. It can be set up as a simple pick of three.
• Add
  Made in Canada, in the banner, on the product pages and in the footer. The wording has rules attached, so it depends on where the ingredients and bottles come from.

Shopify is refusing the PST and GST numbers. For a Quebec business, Shopify asks for a GST number and a QST number (Quebec's version of PST), and each has a set format: the GST number is the 9-digit business number followed by RT and four digits, usually RT0001, and the QST number is 10 digits followed by TQ and four digits, usually TQ0001. A common mix-up is the Quebec enterprise number (NEQ) on incorporation papers, which looks like a tax number but isn't one.

The searches each page should answer. One search per page, so the site's own pages do not compete with each other. How often the site appears for each search, and where it ranks, show up once Google Search Console is connected (Part Three). Until then each term either comes from the live Google checks run on 20 September or is marked as suggested.

Google reads 8 main headings on the homepage and a visitor sees none of them: all 8 sit inside two older banner sliders that no longer show. The headline people do see is not marked as the main heading. Every first-time visitor also lands on the Animals side, while 14 of the 15 orders since 26 September were for the human wash. And the Shop buttons on the homepage go to a page that sends the visitor straight back to the top of the homepage.

The single bottle, the Double Bundle and the 3 bottle bundle share one page layout, which shows the same opening paragraph on all three. The description written for each product in Shopify never appears on the page, and the Description tab holds the dog wash paragraph: "For dogs with lingering odor, greasy coats, dandruff, itchy skin...". The question answers and the Shipping and Sizes boxes are empty.

Worth knowing: the Family Pack's human, human, human option is $79, which is $5.50 less than the 3 bottle bundle before shipping.

Each product page gets its own description, built from the one already written in Shopify, which is a good base. It covers:

• What
  What it is and what is in it: the bottle size and the ingredients in plain words.
• Where
  Where people use it: armpits, groin, skin folds, feet and scalp, for odour, oily skin and flakes.
• How
  How to use it, from the How It Works steps: all over for 2 to 3 minutes then rinse, or on one spot on dry skin for 3 to 5 minutes, then water, rinse and dry well.
• Answers
  Real answers in the four question slots (sensitive skin, how often, fragrance, which product) and the Shipping box.
• Care
  One honest line: if skin is broken, weeping or not improving, see a doctor. STANK is the daily wash, not the treatment.

Dog, Horse and Sphynx each open with their own copy now. The dog paragraph also sits in the Description tab on all of them, the question answers are empty, and the Family Pack has no description at all.

• Horse
  "Try STANK first" appears three times in one paragraph; once is enough, beside the line about seeing the vet. "We've also seen hair grow back" becomes a description of what was photographed.
• Sphynx
  A short how-to for bathing a Sphynx with STANK: how much, how long on the skin, rinsing, drying and keeping warm, how often. Most of what Google shows for this search is how-to content.
• All
  Real answers in the five question slots, starting with "Will it remove strong pet odors?".

The Human and Animal collection pages carry a heading ("Collection: Human") and the product cards, with no words at all, and nothing in the menu, on the homepage or on the product pages links to them. A short introduction on each gives Google something to rank, and the Shop links in the menu and on the homepage then point there.

• How It Works
  Suggested heading: How STANK Works: Clear the Buildup, Reset the Skin Surface. The page has no main heading, and all five question answers are placeholder text. They get real answers: pet odours, sensitive skin (with a patch test), how often, fragrance, and which product. The line "Bacteria & fungi drive about 70% of common skin problems" needs its source on the page; without one, it becomes the homepage's version, "Bacteria & fungi can contribute to unwanted skin concerns." The word microbiome goes in, because it is the accurate and searchable term for what the page describes.
• Reviews
  Suggested heading: STANK Reviews: Real People, Real Pets, Real Smell Tests. No main heading at present. Split into people and pets, with each quote shown once and tester quotes labelled as tester feedback, and the Judge.me reviews joining them as they come in.
• 11 Reasons
  Suggested heading: 11 Reasons STANK Beats Soap for BO and Sweaty Skin Folds (Intertrigo). The page has no main heading and goes straight into Reason 01. A short opening paragraph says who it is for and when to see a doctor. Lines such as "killing them on contact" and "heals and seals these tiny cracks" are rewritten as what the wash does, the "[1.0]" markers get real references or come out, and the note from the foot of the eczema guide goes at the bottom.
• Eczema guide
  The heading stays. Article titles have been pasted into the middle of sentences ("before the itching starts Time."), the labels "The Selling Feature:" and "The Reality:" are still in the text, and the footnotes point to other brands' websites. The five buttons each promise something different and all open the single bottle, so each gets a label that says where it goes. The note at the foot stays.

After the wording above, in this order. Each one answers a search no page on the site answers today.

• Page
  Our Story, from the founder video (section 7).
• Page
  How To Use, for people. The page exists but is empty, and the menu goes to the animal version.
• Page
  Ingredients: the full list, in plain words. There is no ingredient list anywhere on the site.
• Page
  A Sphynx bathing guide with the product inside it, for the how-to searches.
• Page
  A skin folds page for the jock itch and intertrigo searches, building on the 11 Reasons guide.

One campaign is running, LVRG Sales, at $100 a day: one live ad set, Canada, no interests, Meta's Advantage+ audience on, four ads running. The account has recorded 15 purchases since it opened, 12 of them from this campaign.

• Working
  The Meta pixel is installed and recording visits, product views, add to carts, checkouts and purchases.
• How
  Each ad gets about a week before it is judged or changed. An edit sends the ad back through review and can restart Meta's learning.
• How
  The older paused ads still carry the full product set, animal products included, so any that come back on are switched to the human set first.

People who have already shown interest, shown a different ad to bring them back. It is usually the best value campaign an account like this can run, and it needs very little budget because the audience is small.

• Who
  Add to cart and checkout are too few on their own, so the audience combines them with product page visitors, people who viewed several pages, people who watched a quarter or more of a video ad, and people who engaged with the ads or the Facebook and Instagram accounts. The final size is set when the audience is built.
• What
  More information, not a discount. The founder and product explainer videos, and what the wash does for skin and smell. The original Founder Reel, currently paused, is the first choice.
• Budget
  A small share of the $100 a day, moved across from the current campaign so the total spend stays the same.

A second campaign beside the current one, with each ad set built around a group of interests and judged on cost per purchase against the broad campaign. The current campaign keeps running as the benchmark. If broad turns out better, that is worth knowing too.

Should the strongest ads from the September traffic test come back as still images?

In that test the top two ads drew 527 and 437 link clicks, and more than half of the 39 that ran drew 2 or fewer. Every click landed on the old homepage choice screen, so the test shows the message stops people scrolling; it cannot show whether it sells. The strongest, the "blow-dry under my boobs" image, is already back in the current campaign as Boob. There it has had about $30, one add to cart and no purchase yet, because it shares a budget with Skin Won't Heal (the next question). Its fair test is its own ad set, pointed at the Double Bundle, inside the interest test, with BO 5, the second strongest, beside it.

Why is the sebderm ad getting so little spend?

3 October is part of a day. Two paused ads spent a further $1.26.

All four ads share one ad set, and Meta moves the money to the ad it already expects to sell. Skin Won't Heal produced all 4 purchases in those two days, so it gets almost everything, and a new ad in the same ad set rarely gets enough to prove itself. The fix is to give the sebderm ad its own ad set with its own budget, as part of the interest test.

Google Ads reaches people at the moment they search for the problem, which suits this product well. It comes after the shop work. The groundwork goes in now: Google Analytics, the Google tag and Merchant Center through Shopify's own Google app, and Search Console alongside it, so there is data to work from when it starts and the search terms in Part Two can be measured.

• Later
  Landing pages for each skin concern, not in the menu, matched to the ads.
• Later
  A "why are you here?" picker that sends visitors to the right product for their concern.
• Later
  Reddit, within the next six months, once the above is settled.

1. NowRetargeting campaign.
2. NowGoogle Analytics, the Google tag, Merchant Center and Search Console.
3. When the numbers arriveTax registration in Shopify.
4. Once approvedReviews: review requests, the testimonials, the product pages and the cart.
5. Once approvedShipping and returns, the banner and the line under Add to Cart.
6. Once approvedThe loading screen, the logo and the tidy-ups.
7. Page by page, as approvedThe search wording in Part Two.
8. When the video arrivesThe founder story: the page, the product pages and the ads.
9. Once the list is approvedThe interest test.
10. After the shop workGoogle Ads, then Reddit.

Almost everything above is done from this side. Three things need Julie.

• Approve
  This plan, including the changes customers will see in Part One, and the wording in Part Two page by page, with the product facts listed at the end of section 16.
• Send
  The founder video, next week.
• Send
  The GST and QST numbers, by email.