Beyond aesthetics: a social media strategy for lifestyle brands that has to convert

A pretty feed is not a strategy

A lifestyle brand sells a feeling before it sells a cushion. That is true, and it is also the sentence agencies hide behind when the feed looks lovely and the orders do not arrive. A social media strategy for a brand that sells fashion, homeware or food has one job. It has to turn attention into orders at a cost the margin can carry, and it has to do that again next month.

So this article is about the mechanics, not the mood board. Every section ends in a number, because a feed you cannot measure is a hobby.

Three rules from my own style guide sit underneath everything here. Organic content teaches and shows. Ads convert. Retargeting only ever reminds people who have already shown intent, and it never introduces the brand to a stranger. Keep those three apart and most of the confusion in social media reporting disappears.

To carry the numbers I have invented a shop. Long Walk Homeware sells linen, stoneware and candles from a small studio in Galway and ships all over Ireland. Aoife runs it, does the marketing herself, and has about €3,000 a month for paid social. Her average order is €64 and her gross margin is 55%, so an order leaves €35.20 to pay for the marketing that brought it in. Nobody real, and every figure below is made up to show the method.

Know the brand's story before you post anything

Every lifestyle brand is a promise about who the customer becomes. Before anyone picks a platform or writes a caption, that promise has to be written down in one paragraph that a stranger could read back to you. Who buys, what are they trying to get done, and what does the product change about an ordinary Tuesday evening for them.

This is the part most brands skip because it feels soft. It is the hardest part. The platforms and formats change twice a year. The story should not. If you want the formal version of this thinking, the StoryBrand framework and Jobs To Be Done both do the job. The plain version is three questions.

Worked example: Long Walk writes its paragraph

Aoife sat down with her last 200 orders and the notes from 15 customer emails. The pattern was not "people who like linen". It was renters in their late twenties and thirties who cannot change the walls, the floor or the kitchen, and who make a rented flat feel like theirs through the things on the table and the bed. Her paragraph came out like this: "For people who rent and still want a home that feels chosen, Long Walk makes the few things you can take with you, made to last longer than the tenancy."

That paragraph now decides everything. The hook of a video is the rented flat, not the stoneware. The creators she works with rent. The retargeting ad says "take it with you", not "20% off". When a format changes next spring, the paragraph does not.

Pick platforms on evidence, not on what the competitor does

The question is never "which platform is hot". It is "where do the people in my paragraph actually spend time, and what do they do there". For Ireland the best public answer is the Reuters Institute Digital News Report, whose Irish edition is produced each year with the DCU Institute for Future Media, Democracy and Society (FuJo) and Coimisiún na Meán. The 2026 survey asked 2,053 Irish online adults, between the middle of January and the end of February 2026, which networks they had used in the last week for any purpose, and which for news.¹ The Irish report adds that 76% had used a smartphone for news in the past week, which makes the phone the main route even for the most sit down kind of content.² That settles the format argument. Everything you make is watched on a phone, held upright.

The weekly use figures are worth having in front of you when you choose.¹

Network	Used in the last week for any purpose, Irish online adults	Used in the last week for news
WhatsApp	81%	27%
YouTube	65%	26%
Facebook	61%	33%
Instagram	53%	23%
TikTok	29%	15%
X	19%	11%

Two honest limits on that table. It measures weekly use, not time spent, so it says nothing about how long anyone stays. And Pinterest, Snapchat and LinkedIn are not in it at all. I know of no Irish primary survey that publishes their use, so I give no Irish figure for them rather than borrow another country's.

The age shape matters more than the totals, and here the Irish data is thinner still: it is published without an age split per network. The nearest evidence is the same survey across all its 48 markets, which is not Irish and should be read as a direction, not a measurement. There, 53% of 18 to 24 year olds use TikTok weekly for any purpose against 23% of the over 55s, and 67% of 18 to 24 year olds use Instagram against 39% of the over 55s.¹ Both networks lean young, TikTok most of all.

Notice what the table does not say. Pinterest is missing from it, and would look small next to TikTok if it were there. But a minute on Pinterest is spent saving a kitchen table to a board called "flat". That is a buying signal, not entertainment. So the evidence has two layers: where your people are, and what state of mind they are in when they are there.

Worked example: Long Walk chooses three platforms

Aoife's paragraph points at renters aged roughly 25 to 40, skewing female. Reading the figures against it, she picks three. Instagram, because 53% of Irish online adults use it every week, because it leans young, and because it is where her existing customers already tag her. TikTok, because it is the youngest skewed network in the survey, 29% of Irish online adults use it weekly, and it is where a video about a rented flat can travel without a budget. Pinterest, because her customer plans a room there before buying anything. No Irish survey measures Pinterest, so she makes that call on the state of mind alone and treats the quarter's numbers as the test.

She drops Facebook as a place to post, even though 61% of Irish online adults use it weekly, because nothing in the Irish data tells her the people in her paragraph are among them, and her own orders say they are not: of her last 200, only a handful came from a Facebook post. She keeps it only as a placement inside Meta's retargeting, where the audience is defined by behaviour on her site, not by age. She ignores X and LinkedIn entirely. Nothing in her paragraph lives there.

Design the funnel, not the feed

A feed is a list of posts. A funnel is a set of jobs, each with a measure. Both Meta and TikTok build their ad systems around this distinction, so it is worth knowing their words. Meta's campaign objectives are awareness, traffic, engagement, leads, app promotion and sales, and the objective sits on the campaign, with each ad set below it holding one audience and its budget.³ Meta's own guidance is to run one ad set per target audience and several ads inside each set so the system can find the creative that works.⁴ TikTok sorts its objectives into awareness (reach), consideration (traffic, video views, community interaction) and conversion (app promotion, lead generation and sales, which it describes as selling "from your TikTok Shop, website, and app").⁵

The point for a lifestyle brand is simple. If you tell the platform you want video views, it will find people who watch videos. If you tell it you want sales, it will find people who buy. These are different people. A campaign that runs the sales objective on a cold audience with a lovely brand film will spend the budget finding the few people who buy from a brand film, which is almost nobody. Each stage of the funnel gets its own objective, its own content, and its own number.

Here is the loop I build for every lifestyle brand. It runs top to bottom, and the last step feeds the first.

Read it from the top. Organic content does the teaching and carries no budget. The click to the site carries a label so it can be traced. The consent banner splits the world into visits you can see and visits you cannot, which matters enormously for the measuring section below. Only people who behaved like buyers get a paid reminder. And the end of the loop is not the order, it is the lesson, because the lesson is what makes next month less expensive than this one.

Worked example: Long Walk maps content to the loop

Aoife writes one line per stage. Discovery is a 15 second TikTok and Reel each week showing one rented room before and after, with the product as the only change. Consideration is a Pinterest board per room and a weekly carousel on Instagram on how to care for linen, run with no budget at all. Intent is anyone who spends 30 seconds on a product page, views two pages, or adds to basket, which her shop records as events. Conversion is one Meta campaign on the sales objective, with one ad set, shown only to those intent audiences, saying "take it with you". Loyalty is a plain email a month after delivery asking for a photo of the piece in its new home. That photo becomes next month's discovery content, which closes the loop.

The paid budget touches two stages only. A small TikTok and Pinterest spend on the traffic objective to get the first visit, and the Meta sales campaign to remind. Nothing is spent on reach for its own sake.

Keep the story consistent while the content changes

Consistency is not repetition. The paragraph stays fixed, the formats and the subjects rotate. A brand that posts the same product shot in nine colours is consistent and boring. One that chases every trend is lively and forgettable. The middle is a small set of recurring themes from the paragraph, with room each month for one timely piece.

TikTok's own creative guidance for performance ads is useful here and applies just as well to Reels. Shoot upright at 9:16, put the hook in the first six seconds, get the point of the video across in the first three, use sound, and add captions or text so it works with the sound off.⁶ None of that is about aesthetics. It is about the fact that the viewer decides in three seconds whether your rented flat is their rented flat.

Worked example: Long Walk's quarter of themes

Aoife sets three themes for July to September. "The move", for everyone who changes flats over the summer. "The first dinner", for the evening you finally cook for friends in the new place. "Taking it with you", for the things that survive three tenancies. Every video, pin and carousel in the quarter belongs to one of the three. The move theme carries the TikTok spend in July and August because that is when people move. The first dinner theme carries September. The third theme is the retargeting line all quarter. One timely slot a month is left open, and in August she uses it for a short video on a warm spell and linen, because that is what her customers were messaging about.

Measure like a scientist, which means knowing what you cannot see

Three things make social media measurement honest. Most brands have none of them.

The first is labels on every link. A UTM parameter is a short tag on the end of a web address that tells Google Analytics 4 (GA4) where the click came from. Google says you should always set utm_source, utm_medium and utm_campaign, and warns that the values are case sensitive, so "Instagram" and "instagram" become two rows in the report.⁷ I have written the full method in Mastering UTM tracking. For social media the rule is one vocabulary, lowercase, with the platform as the source, paid_social or social as the medium, and the theme as the campaign.

The second is consent. In Ireland, Regulation 5(3) of the ePrivacy Regulations 2011 allows nothing to be stored on, or read from, a visitor's device unless the visitor has consented and been given clear and comprehensive information about the purpose.⁸ The Data Protection Commission (DPC) spells out what that means in its cookie guidance: cookies related to advertising are not strictly necessary and must be consented to, and analytics cookies need consent too, although first party analytics strictly limited to aggregate statistics are unlikely to be an enforcement priority.⁹ In practice a consent banner means your analytics sees only the share of visitors who said yes. That share is not an error. It is the sample you have. I have written up how to build the tracking around it in Event tracking and consent mode. Here the practical rule is: your shop's order count is the truth, the analytics count is the consenting share of it, and you divide the second by the consent rate before you compare channels.

The third is knowing what the platforms count. Meta's reporting offers conversions counted within one, seven or 28 days of a click, within one, seven or 28 days of a view, and within one day of an engaged view, and shows a "default attribution window" value alongside.¹⁰ TikTok defaults pixel and web advertisers to a seven day click and one day view window, and lets you choose one, seven, 14 or 28 days for clicks and off, one or seven days for views.¹¹¹² A view through conversion means someone saw the ad, did not click, and bought within the window. Your analytics tool never sees the view, so the platform's number will always be larger than yours, and neither is lying. They count different things. Pick one yardstick, usually the consent adjusted GA4 count on a seven day click basis, and use it for every platform.

The number I watch most is the cost per engaged visitor, because it tells me what a platform's attention costs before any buying happens.

$$\text{CPEV} = \frac{S}{V_e}$$

Here $\text{CPEV}$ is the cost per engaged visitor, $S$ is the spend on that platform in the period, and $V_e$ is the number of visitors from it who did something that shows interest: stayed 30 seconds, viewed two pages or added to basket. It is a rate, a cost over a count of events, so everything about denominators in Foundations: ratios and rates applies. In particular, never compare it across platforms unless $V_e$ is defined identically on all of them.

Worked example: Long Walk's quarter, in the share it can see

In July to September, 70% of Long Walk's visitors accepted analytics. Aoife spent €5,000 on Meta, €2,500 on TikTok and €1,500 on Pinterest, all with tagged links. GA4 recorded 4,100 engaged visitors from Meta, 2,200 from TikTok and 2,600 from Pinterest.

$$\text{CPEV}_{\text{Meta}} = \frac{5{,}000}{4{,}100} = \text{€}1.22 \qquad \text{CPEV}_{\text{TikTok}} = \frac{2{,}500}{2{,}200} = \text{€}1.14 \qquad \text{CPEV}_{\text{Pinterest}} = \frac{1{,}500}{2{,}600} = \text{€}0.58$$

In each fraction the top is $S$, the spend on that platform, and the bottom is $V_e$, the engaged visitors it sent that GA4 could see. Because the consent rate is the same 70% on every platform, dividing all three by 0.70 would change the euro amounts but not the order, so for ranking platforms the raw figures are grand. For the absolute cost they are not: the true cost per engaged visitor on Pinterest is nearer €0.58 multiplied by 0.70, which is about €0.41, because three in ten of the engaged visitors were never counted.

GA4 recorded 148 orders from Meta, 41 from TikTok and 52 from Pinterest. Meta's own report said 190 and TikTok's said 58. The gap is the view through conversions and the people who said no to the banner. Aoife does not argue with any of these numbers. She writes down which one she will use, and uses it everywhere.

Compare creatives with enough data, which is more than you think

Every lifestyle brand runs two videos and declares a winner after a weekend. Two videos with 40 and 55 clicks have told you nothing. The sample you need to tell two rates apart is a fixed piece of arithmetic, worth doing once so you stop guessing.

$$n = \frac{(z_{\alpha/2} + z_{\beta})^2 \left[p_1(1-p_1) + p_2(1-p_2)\right]}{(p_1 - p_2)^2}$$

Here $n$ is the number of impressions, or visitors, each creative needs. $p_1$ and $p_2$ are the two rates you want to be able to tell apart, for example a click rate of 1.2% against 1.5%. $z_{\alpha/2}$ is 1.96 for a 5% significance level, meaning you accept a one in 20 chance of calling a difference that is not there. $z_{\beta}$ is 0.84 for 80% power, meaning an eight in ten chance of spotting the difference if it is real. The derivation and the trap of "deciding the test length from the calendar" are in Foundations: sample size, and the broader question of how much data a campaign needs has its own article.

Worked example: two Long Walk videos

Aoife wants to know whether "The move" or "The first dinner" gets the better click rate on TikTok. Her videos usually run at about 1.2%, and a lift to 1.5% would be worth knowing about.

$$n = \frac{(1.96 + 0.84)^2 \left[0.012 \times 0.988 + 0.015 \times 0.985\right]}{(0.015 - 0.012)^2} = \frac{7.84 \times 0.02663}{0.000009} \approx 23{,}200$$

Each video needs about 23,200 impressions, which on TikTok is a few days and a few euro. The symbols are as above: $p_1$ is 0.012, $p_2$ is 0.015, and the squared difference on the bottom is tiny, which is why $n$ is large. The formula, and why 80% power is the usual compromise, is in Foundations: sample size. Compare that with testing the same two videos on orders rather than clicks. Her landing page converts about 2.0% of visitors, and she would want to see 2.5%. The same formula with $p_1 = 0.02$ and $p_2 = 0.025$ gives about 13,800 visitors per video. At her Meta cost per click of roughly 54 cent, that is about €7,450 per video, more than her whole Meta budget for the quarter. So she compares creatives on click rate, where she can afford the sample, and then watches the order rate of the winner over the following months rather than pretending a weekend settled it.

Improve targeting before you touch the spend

When a campaign disappoints, the reflex is to cut the budget or double it. Neither is a diagnosis. The diagnosis is three questions. Is the objective right for the stage, or are you asking the sales objective to introduce the brand. Is the audience defined by behaviour or by a guess about demographics. Is the retargeting reminding people who showed intent, or hosing everyone who ever scrolled past.

The third question is where the money goes. A retargeting ad set built on "everyone who visited the site" is paying to remind people who bounced in two seconds. One built on "added to basket in the last seven days" is reminding people who were halfway to buying. Meta's structure makes this easy to do properly because every ad set is one audience with its own budget.⁴ The deeper version of audience building, including lookalikes, is in Lookalike audiences explained, and the case for feeding your own data back into bidding is in Auto bidding on first party data.

Worked example: Long Walk fixes TikTok without spending more

Aoife's TikTok cost per order of around €61 looks like a channel to kill. Before she does, she asks the three questions. The objective was traffic, which is right for discovery. The audience was broad, which is right for discovery. The problem is that TikTok was never meant to close. When she looks at the paths of the orders in her Meta retargeting ad set, which only shows to people who added to basket, 38% of them first arrived from a TikTok click. TikTok is filling the basket audience that Meta then reminds.

So she changes nothing about TikTok's budget and two things about its targeting. She excludes people who bought in the last 90 days, so the spend stops reaching existing customers. And she moves the small share of TikTok budget that had been on a conversion objective back to traffic, because at her volume the sales objective had nothing to learn from. Her TikTok cost per engaged visitor stays at about €1.14, and the Meta retargeting set gets a bigger basket audience for the same money.

Create with people who already live the story

A creator collaboration is a bet, and it should be sized like one. The creator who rents a one bedroom flat in Cork and genuinely owns your candles is worth more to a lifestyle brand than a studio with a lighting rig, and you should expect a wide range of outcomes from each one. The honest way to decide a fee is the expected value: each outcome times its chance, added up.

$$E[X] = \sum_{i} p_i \, x_i$$

Here $E[X]$ is the expected contribution from the collaboration, each $x_i$ is the contribution margin from one possible outcome, and each $p_i$ is the chance of that outcome, with the chances adding to one. The idea, and its one big weakness, that it says nothing about the time you cannot afford the zero, are in Foundations: expected value.

Worked example: one creator, three outcomes

A creator in Limerick with a modest following asks Aoife for €600 plus product worth €90, so the bet costs €690. From her last few collaborations Aoife estimates a 20% chance the video travels and brings 60 orders, a 50% chance of a normal result of 18 orders, and a 30% chance it falls flat at four orders. Each order leaves €35.20.

$$E[X] = 0.20 \times (60 \times 35.20) + 0.50 \times (18 \times 35.20) + 0.30 \times (4 \times 35.20) = 422.40 + 316.80 + 42.24 = \text{€}781.44$$

The symbols are as above: the three $p_i$ are 0.20, 0.50 and 0.30, and the three $x_i$ are €2,112, €633.60 and €140.80. Against a cost of €690 the expected profit is about €91. That is a yes, but a small one, and it comes with two conditions. She never pays a single creator a fee that would hurt if the 30% outcome lands. And the agreement includes the right to use the video in her own ads, which the sum above gives no credit for and which in my experience is where most of the value of a good creator video ends up. Weighting each outcome by its chance is explained in Foundations: expected value.

Long Walk's quarter in one chart

Pulling the quarter together, here is what each euro bought, on the one yardstick Aoife chose: orders GA4 could see, on a seven day click basis, with creator orders counted through their discount codes.

Each bar is a channel. The height is the spend on that channel divided by the orders attributed to it: €1,500 over 52 for Pinterest, €5,000 over 148 for Meta, €2,070 over 54 for the three creators, and €2,500 over 41 for TikTok. The line Aoife holds in her head is €35.20, the contribution an order leaves before marketing. Pinterest and Meta sit under it. The creators sit just over it on first orders, before the ad use of their videos. TikTok sits far over it, and the previous section explained why that is the wrong way to read a discovery channel.

Two adjustments change the picture, and both are the kind of thing a scientist writes down before looking. First, the consent rate. The paid channels were measured through GA4, which saw 70% of visitors, so their true costs per order are about 70% of the bars: roughly €20 for Pinterest, €24 for Meta and €43 for TikTok. The creator figure does not move, because discount codes are counted in the shop, not in analytics. On the adjusted figures the creators are the most expensive channel bar TikTok, which is the opposite of what the raw chart says. Second, the stage. TikTok's job was the first visit, and on cost per engaged visitor it matched Meta. Judged on its own job, it did it.

What Aoife does with this is dull and correct. Pinterest gets another €500 a month, because its cost per order has room to rise before it meets the margin. Meta stays where it is. TikTok stays on traffic and gets judged on engaged visitors. The creators are kept, at three a quarter, each one a bet she can afford to lose.

What to do this week

Write the paragraph. Read it against the Irish weekly use figures, be honest about the networks nobody in Ireland measures, and pick at most three platforms, each with a reason that names a state of mind, not just a reach number. Give every stage of the loop one objective, one kind of content and one number. Tag every link with the same lowercase vocabulary. Write down your consent rate and the one yardstick you will use for every platform, then stop arguing with the platforms' own counts. Work out, once, how many impressions a creative test needs, and never call a winner before that. Build retargeting on intent and nothing else. Size each creator as a bet. Then feed what you learned back into the paragraph, because that is the only part of this that compounds.

Footnotes

1. Reuters Institute for the Study of Journalism, Digital News Report 2026, Ireland country page (pp. 94 to 95) and social media chapter, Reuters Institute, June 2026, https://reutersinstitute.politics.ox.ac.uk/sites/default/files/2026-06/DNR%202026%20FINAL_2.pdf

2. Coimisiún na Meán and DCU FuJo, Digital News Report Ireland 2026, Coimisiún na Meán, June 2026, https://www.cnam.ie/app/uploads/2026/06/Digital_News_Report_2026.pdf

3. Meta, Ad Campaign Group, Marketing API Reference, Meta for Developers, 2026, https://developers.facebook.com/docs/marketing-api/reference/ad-campaign-group/

4. Meta, Marketing API overview: campaign structure, Meta for Developers, 2026, https://developers.facebook.com/docs/marketing-api/campaign-structure

5. TikTok, Choose the Right Objective, TikTok Ads Manager Help Center, 2026, https://ads.tiktok.com/help/article/choose-right-objective?lang=en

6. TikTok, Creative best practices for performance ads, TikTok Ads Manager Help Center, 2026, https://ads.tiktok.com/help/article/creative-best-practices?lang=en

7. Google, URL builders: Collect campaign data with custom URLs, Google Analytics Help, 2026, https://support.google.com/analytics/answer/10917952

8. European Communities (Electronic Communications Networks and Services) (Privacy and Electronic Communications) Regulations 2011, S.I. No. 336 of 2011, Regulation 5, Irish Statute Book, 2011, https://www.irishstatutebook.ie/eli/2011/si/336/made/en/print

9. Data Protection Commission, Guidance note: cookies and other tracking technologies, DPC, April 2020, https://www.dataprotection.ie/sites/default/files/uploads/2020-04/Guidance%20note%20on%20cookies%20and%20other%20tracking%20technologies.pdf

10. Meta, Ads Action Stats, Graph API Reference, Meta for Developers, 2026, https://developers.facebook.com/docs/marketing-api/reference/ads-action-stats/v23.0

11. TikTok, Fuel your TikTok campaign measurement with flexible attribution windows, TikTok For Business Blog, 23 June 2022, https://ads.tiktok.com/business/en-GB/blog/flexible-attribution-windows-campaign-measurement

12. TikTok, How to Create Attribution Windows for your Ad Group, TikTok Ads Manager Help Center, 2026, https://ads.tiktok.com/help/article/attribution-settings-at-the-ad-group-level?lang=en