Creative Diversity Audit
Pott Candles
2026-08-18 · last 30 days (2026-07-19 to 2026-08-18) · concentration on 90 days

The short version

Pott's creative engine is genuinely diverse where most accounts fail: 24 distinct concepts live, no single ad or concept dominating spend, frequency at 1.63, and a launch cadence well above the top quartile for this spend level. The problem is not how much gets made, it is what gets made. Four arguments take 87% of spend (founder and craft story 34%, sensory product desire 24%, refill economics 18%, social proof 11%), while the arguments that would open new audiences have nothing in them. No gifting ad exists ten weeks before Q4. The anti-disposable case that is the whole brand runs £561 across two ads. Press coverage Pott has held since February has never appeared in an ad. And there is not one text-led static, one before-after comparison or one longform explainer in the account. Next month's fix is 35 executions aimed at the empty lanes, not more of the full ones.

£37,411
30-day spend
571
ads with spend
24
live concepts
12
empty lanes
35
executions to build

Scorecard: seven checks

Concepts counted with the two-axis rule: two ads are different only if they differ on at least two of avatar, argument and content type.

GREEN1 · Live concept count24 distinct concepts above £100

Threshold is 6 to 10+ per proposition. Pott runs 24 genuinely different argument-and-content combinations, counted after collapsing scent variants and copy duplicates under the two-axis rule. This is the account's real strength.

GREEN2 · Spend concentrationtop 5 ads = 20.8% of 90-day spend

Red line is 70%. The top five ads take 20.8% of 90-day spend and span four different concept families, so no single execution is load-bearing. The Ad 84 pattern from July, where one ad ate 96% of a campaign, has resolved.

GAP3 · Coverage4 argument lanes empty

Four argument lanes are empty and eight content types have nothing in them. Gifting is the urgent one: zero spend ten weeks before Q4, despite the known December shift toward male gift buyers. Press, objection handling and identity all sit at zero too. This is a supply problem, not a targeting one: the fix is making the missing creative so Meta has it to choose from, never splitting delivery by stage.

WATCH4 · Format mixvideo 54% · static 46% · carousel 0%

Video and static are both inside the 30 to 40 percent guardrails, which is healthy. Carousel is functionally dead at £11 of £37.4k against a 10 to 15 percent guide, and there is no experimental lane at all: no text-led statics, no longform, no humour.

GREEN5 · Style spread13 content types live, top 18%

Thirteen content types carry spend and the top one takes only 18%, which is unusually well spread. The missing styles are the deliberate ones rather than the obvious ones: no press graphic, no comparison visual, no stats card, no explainer.

WATCH6 · Messenger mixcustomer-creator 44% · no-person-product 35% · founder 18% · whitelisted-partner 4%

Three messenger types are genuinely live, which passes. But whitelisted and partnership ads are 3.7% of spend on a single productive ad, while Meta reports creator-handle ads at 19% more clicks and 10% more conversions than brand-handle equivalents. There is no expert voice and no press voice at all.

GREEN7 · Freshness & fatigue307 new ads but 14 genuinely new concept lanes · freq 1.63

Raw cadence is far above the top-quartile bar and fatigue is not a problem: spend-weighted frequency is 1.63 and only three ads with real spend sit above the 2.5 cold-frequency line. The caveat is distinctness. 307 new ads produced 14 genuinely new concept lanes, and the bulk-launched UGC testimonial lane averages £29 of delivery across 153 ads, which is Meta declining almost all of what it was given. That is volume without diversification.

What you have enough of, and what is missing

Every ad is mapped onto a fixed checklist of arguments and content types that should exist, so empty lanes are visible. An empty lane leaves no trace in the data, which is why the checklist is fixed rather than derived from what is already running. Saturated means stop making more of this, not stop spending on it: Meta allocates the delivery, this allocates production hours.

By argument, what the ads say

VerdictLaneSpendShareProductive adsWhat to do about it
SATURATED Founder / craft / origin story
Enough coverage. Refresh winners only, no new sub-variants
£12,808 34.2% 22 Enough coverage. Refresh winners only, no new sub-variants
SATURATED Sensory / product desire
Enough coverage. Refresh winners only, no new sub-variants
£8,803 23.5% 14 Enough coverage. Refresh winners only, no new sub-variants
SATURATED Value / economics
Enough coverage. Refresh winners only, no new sub-variants
£6,715 17.9% 8 Enough coverage. Refresh winners only, no new sub-variants
SATURATED Social proof / testimonial
Enough coverage. Refresh winners only, no new sub-variants
£4,214 11.3% 8 Enough coverage. Refresh winners only, no new sub-variants
HEALTHY Subscription / value ladder
Working. Keep it warm, do not expand
£2,722 7.3% 3 Working. Keep it warm, do not expand
THIN Features & benefits
Under-supplied. Needs more distinct executions
£982 2.6% 1 Under-supplied. Needs more distinct executions
THIN Us vs them / anti-category
Under-supplied. Needs more distinct executions
£561 1.5% 2 Under-supplied. Needs more distinct executions
THIN Problem, solution
Under-supplied. Needs more distinct executions
£482 1.3% 1 Under-supplied. Needs more distinct executions
GAP Identity / ritual / aspiration
Nothing here. Test the lane with distinct executions, not one ad
£124 0.3% Nothing here. Test the lane with distinct executions, not one ad
GAP Gifting / occasion
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
GAP Press / PR / authority
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
GAP Objection handling
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
PARK Education / how it works
Already covered. The refill demo and how-to content lane carries the mechanism job under the value and economics angle.
Absent by choice, no action
Absent by choice, no action
PARK Urgency / offer
By design. Pott runs a launch calendar, not a sale calendar, and discount-led creative would undercut the premium position.
Absent by choice, no action
Absent by choice, no action

By content type, what the ads are

VerdictLaneSpendShareProductive adsWhat to do about it
SATURATED Founder video / talking head
Enough coverage. Refresh winners only, no new sub-variants
£6,663 17.8% 13 Enough coverage. Refresh winners only, no new sub-variants
SATURATED Unboxing
Enough coverage. Refresh winners only, no new sub-variants
£6,661 17.8% 8 Enough coverage. Refresh winners only, no new sub-variants
SATURATED How-to / tutorial
Enough coverage. Refresh winners only, no new sub-variants
£5,650 15.1% 9 Enough coverage. Refresh winners only, no new sub-variants
SATURATED UGC testimonial
Enough coverage. Refresh winners only, no new sub-variants
£4,438 11.9% 9 Enough coverage. Refresh winners only, no new sub-variants
HEALTHY UGC demo / first use
Working. Keep it warm, do not expand
£5,288 14.1% 3 Working. Keep it warm, do not expand
HEALTHY Lifestyle static
Working. Keep it warm, do not expand
£2,981 8% 7 Working. Keep it warm, do not expand
HEALTHY Product / studio static
Working. Keep it warm, do not expand
£2,368 6.3% 5 Working. Keep it warm, do not expand
THIN Social-native / story style
Under-supplied. Needs more distinct executions
£1,505 4% 3 Under-supplied. Needs more distinct executions
THIN Whitelisted / partnership ad
Under-supplied. Needs more distinct executions
£1,401 3.7% 1 Under-supplied. Needs more distinct executions
THIN Behind the scenes / making-of
Under-supplied. Needs more distinct executions
£377 1% 1 Under-supplied. Needs more distinct executions
GAP Review / press screenshot graphic
Nothing here. Test the lane with distinct executions, not one ad
£38 0.1% Nothing here. Test the lane with distinct executions, not one ad
GAP Problem-solution video
Nothing here. Test the lane with distinct executions, not one ad
£29 0.1% Nothing here. Test the lane with distinct executions, not one ad
GAP Carousel (lineup / steps / styling)
Nothing here. Test the lane with distinct executions, not one ad
£11 0% Nothing here. Test the lane with distinct executions, not one ad
GAP Text-led / listicle static
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
GAP Before / after / side-by-side
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
GAP Stats / facts card
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
GAP Longform explainer / VSL
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
GAP Skit / humour / relatable
Nothing here. Test the lane with distinct executions, not one ad
Nothing here. Test the lane with distinct executions, not one ad
PARK B-roll + voiceover
Already covered inside the founder voiceover lane.
Absent by choice, no action
Absent by choice, no action
PARK Brand film / hero manifesto
Deprioritised. Hero brand films are expensive and rarely earn direct-response delivery at this spend level.
Absent by choice, no action
Absent by choice, no action
PARK Podcast / interview style
Deprioritised. Needs a recurring host format Pott does not have; revisit if the founder starts appearing on podcasts.
Absent by choice, no action
Absent by choice, no action
PARK Green screen / reaction
Off-brand. Reaction and duet formats fight the calm, crafted register the brand sells on.
Absent by choice, no action
Absent by choice, no action
PARK Meme / lo-fi native
Partly covered by the social-native lane, and the full meme register is off-brand for a premium ceramic product.
Absent by choice, no action
Absent by choice, no action
PARK Other / unclassified
Not a real lane, it is the unclassified bucket.
Absent by choice, no action
Absent by choice, no action

Productive ads = executions that earned £100+ of delivery in the period. The gap between total ads and productive ads is the tell: a lane holding many ads but few productive ones is Meta declining most of what it was given.

Spend distribution

By argument

Founder / craft / origin story
£12,80834.2%
Sensory / product desire
£8,80323.5%
Value / economics
£6,71517.9%
Social proof / testimonial
£4,21411.3%
Subscription / value ladder
£2,7227.3%
Features & benefits
£9822.6%
Us vs them / anti-category
£5611.5%
Problem, solution
£4821.3%
Identity / ritual / aspiration
£1240.3%

By content type

Founder video / talking head
£6,66317.8%
Unboxing
£6,66117.8%
How-to / tutorial
£5,65015.1%
UGC demo / first use
£5,28814.1%
UGC testimonial
£4,43811.9%
Lifestyle static
£2,9818%
Product / studio static
£2,3686.3%
Social-native / story style
£1,5054%
Whitelisted / partnership ad
£1,4013.7%
Behind the scenes / making-of
£3771%
Review / press screenshot graphic
£380.1%
Problem-solution video
£290.1%
Carousel (lineup / steps / styling)
£110%

Live concepts

ArgumentContent typeSpendAdsROAS
Founder / craft / origin storyFounder video / talking head£6,663821.87
Sensory / product desireUnboxing£4,79482.12
Value / economicsHow-to / tutorial£4,447342.09
Social proof / testimonialUGC testimonial£4,1111531.80
Founder / craft / origin storyUGC demo / first use£3,049102.36
Sensory / product desireProduct / studio static£2,186231.66
Subscription / value ladderUGC demo / first use£1,98321.25
Value / economicsWhitelisted / partnership ad£1,32510.91
Sensory / product desireSocial-native / story style£1,027111.12
Founder / craft / origin storyUnboxing£1,024100.91
Features & benefitsLifestyle static£98211.56
Founder / craft / origin storyHow-to / tutorial£969421.91
Value / economicsUnboxing£842161.48
Sensory / product desireLifestyle static£705462.36
Subscription / value ladderLifestyle static£60084.94
Problem, solutionLifestyle static£48262.50
Founder / craft / origin storyBehind the scenes / making-of£377131.25
Founder / craft / origin storySocial-native / story style£339200.35
Us vs them / anti-categoryUGC testimonial£32710.14
Us vs them / anti-categoryHow-to / tutorial£22980.19
Founder / craft / origin storyLifestyle static£213170.90
Founder / craft / origin storyProduct / studio static£161220.46
Subscription / value ladderSocial-native / story style£13922.77
Identity / ritual / aspirationUGC demo / first use£11030.81

Next month's build plan

35 distinct executions across 16 briefs. Each brief is an argument crossed with a content type, so one shoot closes a gap on both axes at once. Counted as cells rather than per-axis, because adding the axes separately double-counts every brief that fills two gaps.

Priority 1 16 executions

Gifting / occasion × Unboxing

Gift-worthy unboxing UGC. The reveal of the starter pack as a present being opened, not a purchase being used. Frame the refill as the part of the gift that keeps arriving. Unboxing is already a proven lane here at 17.8% of spend and 2.12 ROAS, so this borrows a working format for an audience that currently sees nothing.

Gifting / occasion × Lifestyle static

Occasion statics for the gifter. Styled gift moment, legible at thumbnail size, one line of copy naming the occasion. Weight male gift-buyer casting for November and December per the gifting audience shift analysis, and lean F25-34 in November.

Gifting / occasion × Carousel (lineup / steps / styling)

Gift guide carousel. Starter pack options by price or by scent personality. Doubles as the first working carousel in an account where the format is currently dead at £11.

Us vs them / anti-category × Before / after / side-by-side

The drawer of dead candle jars versus one Pott that stays. Side-by-side static, no founder, no talking. This is the brand's whole argument and it currently runs £561 across two ads with no comparison visual anywhere in the account.

Us vs them / anti-category × Longform explainer / VSL

Longform explainer walking the refill case end to end: hook, the disposable-candle problem, the mechanism, proof, offer. The format Andrew Faris reports as his most consistent performer, and Pott has none. Expensive, so one well-made attempt rather than a batch.

Value / economics × Text-led / listicle static

Text-led statics carrying the refill economics. Motion measures text-based ads at the highest hit rate of any format at 11.6% and the account runs exactly zero. Cheapest possible test of a proven-productive format, and the refill maths is the argument best suited to words on a page.

Social proof / testimonial × Whitelisted / partnership ad

Creator-handle versions of already-proven UGC. Lowest-lift item on this list: the creative already exists, only the delivery identity changes. Whitelisted spend is 3.7% carried by one productive ad; target 10% or more.

Priority 2 11 executions

Press / PR / authority × Review / press screenshot graphic

Press authority statics. Pott has held an Ideal Home placement since February that has never appeared in a single ad, and the press argument sits at zero spend. Screenshot-style treatment, publication mark visible, one pulled line.

Objection handling × Founder video / talking head

Founder answering the price objection directly: what you actually pay per burn once you refill, and why the vessel costs what it does. Founder video is the single most productive content lane in the account at 13 productive ads, so this points a proven format at an argument that has nothing.

Features & benefits × Text-led / listicle static

Feature callouts: burn time, wax, three wick, made in Stoke. One ad currently carries this entire argument at £982. Pairs with the text-led format above so a single design session serves both gaps.

Value / economics × Stats / facts card

The refill economics as numbers on screen: jars saved per year, cost per burn versus a disposable premium candle. Verify the real figures before build. Refill economics is already a saturated argument, but it has never been expressed as data.

Problem, solution × UGC demo / first use

Customer-shot problem framing beyond cooking smells: the guilt of binning a £40 jar, the candle that dies mid dinner party. The whole problem-solution argument is £482 on one productive ad.

Priority 3 4 executions

Identity / ritual / aspiration × Lifestyle static

The evening ritual: slow, atmospheric, no product pitch until late. Premium candle brands live in this register and Pott spends £124 a month there. Lifestyle static is already healthy, so this is an argument extension rather than a new format.

Sensory / product desire × Carousel (lineup / steps / styling)

Scent lineup or one Pott four rooms styling carousel. Revives a format that is functionally dead in the account.

Sensory / product desire × Skit / humour / relatable

One relatable-humour test. Motion finds humour in only 14% of social ads but 25% of ads from £1M+ spenders. A single test kept on-brand, not a commitment to the register.

Refresh only 4 executions

Refresh the four saturated lanes

Fresh executions of the proven arguments only: founder and craft story, unboxing, refill how-to, UGC testimonial. Refresh to fight fatigue, do not expand into new sub-variants. These four already take 87% of spend, and the UGC testimonial lane in particular is being declined by Meta at 153 ads for £29 of delivery each, so the answer there is fewer and better, not more.

Capacity check: 35 executions a month against a benchmark median of ~18 and top quartile of ~35 for accounts in the $10k-50k/mo band. Absent lanes get 2 to 4 shots rather than 1 because at a roughly 5% winner rate a single ad cannot tell you whether the lane failed or the execution did.

Method

Data: Meta API ad-level insights, 571 ads with spend from 2026-07-19 to 2026-08-18 (£37,411), concentration measured over 90 days. Ads tagged on six dimensions (avatar, awareness, argument, format, content type, messenger) parsed from the naming convention plus ad copy, then collapsed to concepts with the two-axis rule: two ads count as different only if they differ on at least two of avatar, argument and content type.

Thresholds per the Kova creative diversity framework: Meta-official where published (fatigue definitions, partnership-ad lift, the consolidate-structure-diversify-creative position), practitioner benchmarks elsewhere (70% concentration red line, 6 to 10 concept floor, launch-cadence bands, hit rates by format). Meta has never published a video to static ratio; the widely quoted 60-66% video figure is a consumption stat, not creative guidance.

Nothing in this audit recommends splitting delivery by funnel stage, awareness level or audience. Meta decides who sees which ad; the only lever here is what gets made, and therefore what it has to choose from. ROAS is shown for context only, and cold openers should be judged on first-click before acting on any row.