Sell the travel pack after the buy button, and test Dose for your Skin against it

Dose sells a 2oz 24-Pack. Almost nobody buys it: 124 one-time orders in three days, against 915 for the Liver 3-Pack. It is not a bad product. It is a product nobody meets, hidden behind a 16oz / 2oz toggle most shoppers never touch.

Meanwhile the reviews keep raising the exact problem it solves. Customers love the ritual and then hit a wall the moment they leave the house. One of them says it plainly: "I can't figure out how to travel with it, so my consumption will be interrupted a bit for the next several weeks." A customer who breaks the habit stops renewing.

A post-purchase offer costs nothing to test. The order is already banked, and the redesign does not touch this surface. Today the confirmation page says thank you and stops; no post-purchase offer is running at all. If they decline, Dose keeps the sale and we learn.

Because nothing runs today, there is no point spending a variant on an empty page. The control is the simplest offer: the travel pack on its own. V1 keeps the travel pack and adds Dose for your Skin if it is declined. V2 offers Skin on its own. One question, three answers: which offer earns most per order, and is a second chance worth asking for.

Run all three checkout blocks instead of choosing between them

DD05 has been live in the checkout since 26 August. It is the first checkout test Dose has ever run, and it tests three reassurance blocks one at a time against a bare checkout: the brand promise, the customer voice, and the clinical proof. Each sits in its own slot on the page.

The early pattern is not that one of them wins. At the stage these blocks can actually act on, checkout completion, all three sit about 1.5% above the bare control and level with each other: 69.18%, 69.22% and 69.16% against 68.15%. Five days is not a decision and we are not treating it as one. But if the full read still looks like this, then "pick the winner" is the wrong follow-up question.

So the follow-up asks whether they add up. The winning block gets deployed and becomes the new control, exactly as DD02's V2 is being deployed so the redesign is read against it. V1 carries two blocks, V2 carries all three, spread down the page in the slots they already use rather than piled into one. The risk is the reason to test it rather than just ship it: three reassurance blocks on one page can start to read as noise, and each one adds length on mobile, which is 90.7% of traffic.

Make the price per serving the number that leads

Per-serving pricing already exists on some Dose landing pages and not others, and nowhere is it the number that leads. The buy box leads with $90.00 struck through and $63.00. The hypothesis is hierarchy, not price: make the per-serving figure the primary anchor and keep the billed price plainly readable underneath.

There is more than one way to lead with that number, so this runs as several variants rather than one. The lightest changes the buy box and nothing else. A heavier one carries the framing into the imagery around it, so the per-serving figure belongs to the page rather than sitting in a price row on its own. Each variant moves one thing: the copy angle, or how much of the page carries it.

Lead with what customers rave about, then tell them when to expect results

Taste is the most common thing Dose customers say. We counted every review published in the last 30 days: 139 of 395, 35%, mention it. It was 35% in July too, so this is the steady state, not a good month. The PDP already owns the asset with a "Tastes like Orange Juice" tile. The landing page hero does not lead with it.

Underneath it runs the opposite theme. 12 of those 395 volunteer some version of "too early to tell" without being asked, and behind them sits the bigger number: 73.4% of buyers are here for something they cannot observe day to day, a health marker (40.6%) or general prevention (32.8%).

So the test isolates both halves and then combines them. V1 leads the hero with taste, the thing that makes the habit stick. V2 leaves the hero alone and adds a plain expectations timeline, so the wait has a shape. V3 runs both. Read this way, we learn whether a win comes from the appetite, the reassurance, or only from having both. Variants shown are illustrative; the real build gets specced if this is approved.

Put the second system in front of a single-product cart

The redesign is changing the cart, so this is a hold. It is here because it should be ready the day the surface clears. Today the drawer carries line items, a discount field, a total and a button. There is no cross-sell and no upgrade in it at all.

Customers take both systems as one decision, never as two. Two product families land in the same web basket 2.1% of the time. The bundle, sold as one thing, lands in 9.9% of web orders. Packaging beats assembling, which is why the upgrade leads.

So put it in front of them two ways and let the cart tell us which it prefers. V1 upgrades the Liver 3-Pack to the Liver & Cholesterol Bundle 6-Pack, a +$49 step from $63 and the same price as a Liver 6-Pack. V2 simply adds the second product alongside what is already there. Variants shown are illustrative; the real build gets specced if this is approved.

Keep the offer, change the game: an interactive popup

Dose's popup states its offer flat: "UNLOCK UP TO $48 OFF + FREE GIFTS", an image, an email field, a button. Every test so far has reworded that sentence. The last round came back flat to slightly down.

IM8 does something else, and it is not alone: Grüns and other supplement brands are running gamified capture too. IM8's popup is a card-match game: "Try your luck. Match 3 cards to see what you win." The cards all land on 30% off, and only then does the email field appear. The prize never varies. The game does.

So we hold the offer at exactly what Dose gives everyone today and change only the wrapper. Any lift is engagement, not a bigger discount. A card reveal, a scratch-to-reveal, or a goal picker that captures the wellness answer while it reveals. No wheel, no odds, no invented scarcity. Mechanics shown are illustrative; the real build gets specced if this is approved.

Offer the second system at checkout, matched to the cart

Dose sells systems that overlap, and its customers have overlapping problems. The basket does not. 4,535 of 4,682 web orders in three days held exactly one paid product. Two product families appear together on 2.1%.

Ask the buyers and it inverts. 30.5% of those surveyed say they are managing cholesterol too. Intelligems now supports conditional checkout upsells on cart rules, so the offer matches the cart: cholesterol to a liver cart, liver to a cholesterol cart. One product. Not a carousel.

The reach is limited: we can only target one-time orders and only sell one-time items, roughly 10 to 20% of checkout orders. It is a spare for one reason only: the checkout has a queue. Checkout is untouched by the redesign, and this is still the first AOV lever Dose would ever put in front of a buyer at the pay button.