---
description: Oaken &amp; Pike recap of the eight-week Slow Mornings campaign for Marlow Coffee: results, spend by channel, what worked, what did not, and next bets.
---

 

## The headline result

+42%

click-through against the spring campaign.

3.1% of the people who saw the work clicked, against 2.2% in March–April. The campaign came in €2,000 under its €120,000 budget, reached 2.4 million people and brought 8,310 new customers at €14.20 each, 22% cheaper than in spring.

## The numbers that matter

Spend €118k of the €120k budget · 98% used 

Reach 2.4M people · 1.9M in spring · ▲up 26% 

Click-through 3.1% 2.2% in spring · ▲up 42% 

Cost per customer €14.20 €18.20 in spring · ▼down 22% 

New customers 8,310 6,040 in spring · ▲up 38% 

Repeat purchase 40% within 30 days · 31% in spring · ▲up 9 pt 

## Week by week

New customers per week, first orders attributed as described under the channel table. The dashed line is the weekly average of the spring campaign, 755 new customers a week. From week 2 the autumn campaign never dropped below it.

Spring weekly average, 755 new customers 2,000 1,500 1,000 500 0 640 790 880 950 1,010 1,870 1,160 1,010 W1 W2 W3 W4 W5 W6 W7 W8 1 Sep 8 Sep 15 Sep 22 Sep 29 Sep 6 Oct 13 Oct 20 Oct Creator post, 7 Oct, plus €6k moved behind it. 22% of all new customers. 

**Figure 1 · New customers per week against the spring weekly average.** Week 6, highlighted, brought 1,870 new customers, 22% of the campaign total. One of the six creators posted an unscripted follow-up about her morning routine on Wednesday 7 October; it reached 640,000 people in four days and we moved €6,000 of paid social behind it within 48 hours. Week 1 is the only week below the spring line.

## Where the money went

Five channels, one currency: the cost of a new customer. Creators were the cheapest paid customer of the campaign. Out-of-home took 17% of the spend and delivered 5% of the customers.

__Spend and new customers by channel, weeks 1 to 8__
| Channel     | Spend    | New customers | Share of customers | Cost per customer | Cost vs spring |
| ----------- | -------- | ------------- | ------------------ | ----------------- | -------------- |
| Paid social | €52,000  | 3,900         | 47%                | €13.33            | ▼down 26%      |
| Search      | €24,000  | 1,850         | 22%                | €12.97            | ▼down 8%       |
| Creator     | €18,000  | 1,540         | 19%                | €11.69            | ▼down 40%      |
| Email       | €4,000   | 620           | 7%                 | €6.45             | ▼down 9%       |
| Out-of-home | €20,000  | 400           | 5%                 | €50.00            | ▲up 8%         |
| Total       | €118,000 | 8,310         | 100%               | €14.20            | ▼down 22%      |

Attribution: last paid touch within seven days for paid social and search; creator codes and links for creator; order-level for email, counting first orders from subscribers who had not bought before; post-purchase survey for out-of-home. Cost vs spring compares with the March–April 2026 campaign, same method.

## What worked

1. ### Creators beat every paid channel on cost  
Six creators, €18,000, 1,540 new customers at €11.69 each. The week-6 video was not in the brief: the creator posted a second, unscripted piece about her morning routine and it reached 640,000 people in four days. The two best-performing creator pieces of the campaign were both outside the shot list we gave them.
2. ### Moving the budget every Monday  
We reviewed spend weekly and moved €7,000 between lines, out of out-of-home and brand search and into paid social and creators, plus a further €9,000 inside paid social from the “origin” creative set (2.4% click-through) to the “ritual” set (4.2%). Click-through stood at 2.6% at the end of week 2 and finished at 3.1%; the reallocation is most of the difference.
3. ### Email did the repeat work  
The three-send “second bag” sequence, sent 5, 12 and 20 days after a first order, cost €4,000 and lifted repeat purchase within 30 days from 31% to 40%. It is the reason the +9 pt sits next to the acquisition numbers rather than behind them.

## What did not

1. ### Out-of-home did not pay back  
€20,000 across three cities in two flights; we cancelled the third. By post-purchase survey, 400 new customers named the posters as where they first saw Marlow: €50 per customer, 3.5 times the blended figure, on the weakest attribution in the table. We do not recommend out-of-home again at this scale, and bet 3 below is the small test that would settle it.
2. ### Week 1 underperformed, and that is on us  
640 new customers, below the spring weekly average of 755\. The landing page shipped three days late and the first two creatives went live without a test; the “origin” set opened at 2.1% click-through. Our estimate is around 300 customers lost, most of them in the first four days.

## What we learned

* **Creative choice moved results more than bidding.** Between the two creative sets, click-through differed by 1.8 pt. No bid or audience change during the eight weeks moved it by more than 0.3 pt.
* **Brief the ritual, not the shot list.** Both breakout creator pieces were unscripted. The winter brief should describe the morning we want people to recognise and leave the framing to the creator.
* **The offer mattered less than the hour.** “First bag €9” and “free shipping on the first three bags” performed within 4% of each other. Posts published between 06:00 and 08:00 outperformed the same creative published after 17:00 by 31% on click-through.
* **Attribution is now the weakest part of the setup.** Paid social, search and email numbers are defensible. Creator and out-of-home rest on codes and a one-question survey that 4,900 of the 8,310 new customers answered. Bet 3 includes a tracked offer for that reason.

## Next bets

Three proposals for the winter campaign, each with a cost, the effect we expect, and the rule we would use to stop it.

__Proposed bets for winter 2026 to 2027__
| Bet                                                                                                | Cost    | Expected effect                                                                                 | Decision rule                                                        |
| -------------------------------------------------------------------------------------------------- | ------- | ----------------------------------------------------------------------------------------------- | -------------------------------------------------------------------- |
| 1 · Twelve creators for winter, up from six, briefed on the ritual                                 | €36,000 | About 3,000 new customers at €13 or less each, and two more posts of week-6 scale               | Continue only if cost per customer is under €15 at the end of week 4 |
| 2 · Always-on “second bag” sequence, email plus SMS                                                | €6,000  | Repeat purchase within 30 days from 40% to 44%; cost covers the build and three months of sends | Drop the SMS leg if unsubscribes exceed 1.5% on any send             |
| 3 · One out-of-home format only: station panels near the two Marlow cafés, with a QR-tracked offer | €8,000  | A clean read on whether out-of-home can reach €25 per customer, with attribution we can defend  | Remove out-of-home from 2027 plans if it lands above €25             |

## Budget against spend

€118,000 spent of €120,000\. The variances are the Monday reallocations described above; contingency covered €1,000 of the creator top-up and the rest was not needed.

__Budget and spend by line, media and creator fees excluding VAT__
| Line        | Budget   | Spend    | Variance | Why                                                                                               |
| ----------- | -------- | -------- | -------- | ------------------------------------------------------------------------------------------------- |
| Paid social | €48,000  | €52,000  | +€4,000  | Reallocated from out-of-home in weeks 5 to 8, behind the “ritual” set and the week-6 creator post |
| Search      | €26,000  | €24,000  | −€2,000  | Brand terms capped from week 4; 70% of brand clicks came from people who had already bought       |
| Creator     | €15,000  | €18,000  | +€3,000  | Two creators added in week 3, and usage rights for the week-6 video                               |
| Email       | €4,000   | €4,000   | €0       | As planned                                                                                        |
| Out-of-home | €24,000  | €20,000  | −€4,000  | Third flight in week 7 cancelled                                                                  |
| Contingency | €3,000   | €0       | −€3,000  | €1,000 released to the creator line; €2,000 unspent                                               |
| Total       | €120,000 | €118,000 | −€2,000  | 98.3% of budget spent                                                                             |

## Notes on the numbers

* **Window.** 1 September – 26 October 2026, with a seven-day attribution tail to 2 November. Spring comparison: 2 March – 26 April 2026, same method.
* **Click-through.** 189,000 clicks on 6.1 million paid impressions across paid social, search, creator and email. Spring: 103,000 clicks on 4.7 million impressions.
* **Reach.** De-duplicated people across the four digital channels, from platform reporting. The out-of-home audience is estimated by the media owner and is not included.
* **New customers.** First orders in Marlow’s order export, matched to a channel as described under the channel table.
* **Repeat purchase.** Share of new customers who ordered again within 30 days. Measured on weeks 1 to 4 customers (3,260) against the same window in spring; weeks 5 to 8 will be added in the December update.
* **Money.** Media and creator fees, excluding VAT. The Oaken & Pike retainer is billed separately and is not part of the €118,000.