.png)
I started running ads when Facebook ads were a magic potion.
No matter what you did, ads performed. It was 2016; the word “D2C” was sparkling new, and the “fear” connected with meta-advertising was non-existent. We lived in times when profitability through ads was a sure thing if you knew what you were doing.
It is because of the successful brands of this time, that scaled without retention; that scaled without community or socials - the ecosystem got into the bad habit of thinking that meta ads alone can be the solution to all your problems. Kids started dropping out of college starting drop-shipping stores believing that if they could run meta ads, they could build a business. Dropshipping was hot, Alibaba was legal, and raising funds by using the word “D2C” was pretty easy.
The goal of this writeup is to make sure that if you’re a D2C founder from a non-marketing background, I give you behind the scenes understanding of how media buying (running ads) works in 2025; and the reasons for these mechanics with some historical context. It also helps that I will be writing this article as a D2C founder who understands marketing instead of a marketer who understands business.
Cookies were invented in 1994 by Lou Montulli, a Netscape engineer, as a way to store user data for websites—things like login status, shopping cart items, and preferences. Soon, advertisers realized cookies could track users across the web, enabling behavioral targeting, attribution tracking, and remarketing.
For Facebook Ads, cookies became essential in tracking users' interactions beyond the platform, linking ad impressions and clicks to actual conversions.
Why Cookies Are the Foundation of Facebook Ads Tracking
Without cookies, modern ad tracking as we know it wouldn’t exist. Here’s why:
The death of third-party cookies means platforms like Facebook must rely on server-side tracking (CAPI), first-party data, and AI-driven attribution models (like modeled conversions) to maintain ad effectiveness.
Yes you read that right - cookies are dying - lobbyists and privacy activists have won against advertisers, and as a result browsers (which have conventionally stored all our data) are no longer allowed to store this data. This means that there is a need for advertisers and brands to learn how to store and track this data without depending on help from browsers.
This is going to be the largest change in modern advertising since the IOS14 update of 2021.
P.S. the average conversions & roas of brands fell by 40-60% post the IOS update. Even the ones that were the best prepared for such changes fell by 20%.
Third-party cookies—the backbone of digital advertising for decades—are being phased out.
This shift is driven by increasing privacy concerns, stricter regulations (GDPR, CCPA), and the push towards a cookieless web where users have more control over their data.
Third-party cookies allowed advertisers to:
✅ Track users across different websites
✅ Attribute conversions to specific ads
✅ Retarget users who didn’t convert
✅ Build lookalike audiences based on browsing behavior
Without them, platforms like Meta and Google Ads lose visibility on user journeys, making it harder to optimize campaigns and measure ROAS accurately.
Facebook’s ad system heavily relied on third-party cookies for cross-site tracking. When Apple’s iOS 14.5 update (App Tracking Transparency - ATT) forced apps to get explicit user consent, ad targeting and attribution took a massive hit.
With third-party cookies dying, expect:
⚠️ Less accurate tracking for ad conversions
⚠️ Weaker retargeting capabilities
⚠️ Increased reliance on AI-driven attribution
If you have started advertising in late 2024 or early 2025, you must know that this is the most volatile and uncertain time in meta advertising in decades. A shift like this happens every 3-4 years and right now we’re in the middle of a major transition.
Third party data - data stored by a browser
Second party data - data stored by another company (meta or google)
First party data - data stored by a brand/ advertiser (you)
Before we move into learning more about first party data, I want to share the concept of business inflation with you. It’s a simple concept which simply implies that the cost of running your business is getting higher with every passing year. The % of business inflation which used to take decades now is happening at shorter intervals now. Here are some examples:
Key takeaway: Keep increasing your prices & AOV by 15-20% per year.
Before: a facebook media buyer + someone who could design stuff on Canva
Today: a copywriter + facebook media buyer + designer + creative strategist
Before: 10 creatives per month for every 10 lacs we spend
Today: 20-30 creatives per month for every 10 lacs we spend
Before: video ads were not mandatory - maybe 10-20%
Today: 50-60% ads should be video ads
There are several other examples too, mostly outside facebook ads media buying. This has more to do with our ecosystem and economy but I am just pointing out that meta ads is a prime example of this phenomenon - and you should embrace it wholly because it’s not going away.
The solution is not to resist this phenomenon but to be hyper aware of it so that you understand where to invest in to get the maximum ROI.
The onus now is on YOU to invest in your tracking infrastructure. You can no longer be reliant on google, meta and apple to do your bidding for them. A tool like this will ensure that your data is being captured the way it should be and the right quality and quantity of events are being passed to your meta pixel so that the algo can do its job and get you more high quality users.
Cost of setup: Approx $100 for every 10,000 people on your website. Minimum $100/m.
Recommendations
By the time 2025 ends, any brand not using this will be digging their own grave. Infact, i suspect that this could be a mandate before Q2 ends. The earlier we move to first party data tracking the better it will be for us.
You will often hear things like - “targeting is dead” or “creative is the new targeting”....but what do these things actually mean?
In simple words, audiences and interests were one of the main reasons why meta ads were so successful when they started. Setting up interests manually used to take up 60-70% time of advanced media buyers back in the day. Today…not so much.
Here’s why:
The main reason for this is that due to the redaction of third-party cookies and other privacy reasons, the accuracy and effectiveness of these targeting options have drastically fallen. Earlier, if I was married, there was a 100% chance that Meta knew I was married. Now, even if they do know, they are not allowed to share it with an advertiser.
This infamous facebook scandal was another key event that took place during the US elections of 2016.
This lead to thousands of interests and targeting options suddenly disappearing from meta targeting dashboard almost overnight. Important demographic details like political orientation, income, & other such details were redacted from Meta.
Since then, every few months, meta keeps removing important targeting options from the platform.
As of the time of writing, Facebook no longer relies on your targeting selection to find your customers.
It’s now 100% your ad creative — the image or video, the words, and the engagements.
Basically, the algorithm tries to understand your ad creative and decide who to show it to, then optimize the process with feedback from data.
So, if you’re doing the right thing, your ads should get to the right people without you ever having to bother about targeting.
Tik-tok’s method of targeting users was revolutionary. Instead of classifying users into an interest group based on their general behaviour and demographics, tik-tok classifies them as an interest group based on current consumption patterns.
Before: if my interest group says “vegan” - meta will show ads to 100 people who are vegan.
Today: if my interest group says “vegan” - meta will show ads to 100 people are currently interested in topics related to veganism.
This means that targeting is way more dynamic now.
Remember: “interest group” being vegan does not matter anymore. Your creative needs to say or imply vegan for me to signal to meta what my interest group is.
PS. the tik-tok algo has been copied by all platforms now. Without exception, all social platforms including X and Linkedin are also using the same algorithm.
What does this mean for founders?
It means that you don’t have to worry about the “targeting” options set on the meta dashboard anymore. Sure, age, gender, city will always matter - but outside of that, just focus on creatives and you’ll be good.
Final thoughts on data
For meta, everything is data. The photographs on your PDP, the number of reviews on the website, the content in ads, the structure of the landing page, the influencers you work with, the copy on your product pages, the engagement on your social media. All of these things combine to signal to your meta pixel about who your exact ICP is. Keep everything consistent, and focus on creating an abundance of this right type of data, and your ads efforts will be well rewarded.
First-party tools like customer labs & Aimerce take this up a notch and help you create even MORE data points to feed into your meta dashboard. Brands using these tools see upto 40% more events being passed into their meta events manager, leading to anything between a 20% to a 60% improvement in CAC and roas numbers.
CAT framework
As a marketing company, we predicted this switch 2 years ago which is why 60% of our team size today is content & creative optimisation specialists. Exactly 2 years ago, the whole content and creative vertical at Porcellia was outsourced and the only thing we managed inhouse was the technical optimisation of the dashboard.
Importance from 2015-18: 20-30%
Importance from 2018-21: 30-40%
Importance today: 60-70%
This has always been important & rewarded well - now it’s a necessity.
Easy-to-remember formula
Better content will directly reduce CAC
Better tech will increase your AOV & conversion rates
If facebook ads results became predictable(consistent), most founders would report a 60-70% in their day-to-day anxiety about running their business. THAT is how insanely bad meta is at delivering predictable returns for your business.
Below I am sharing some common issues with Meta that you should be aware of before starting this journey. The goal is to help you understand that most likely what you’ll go through is common, and there is no need to panic. (in most cases)
You should be tracking meta performance weekly or monthly. Definitely not daily. Yes your marketing team obv. should be looking at everything daily but as a founder, looking at the inconsistencies in day-to-day performance may overwhelm you.
A creative that gets you 3 roas consistently for 3 months is better than a creative that gets you 6 roas for 1 week and 1 roas the next week. Study your most consistent creatives - not your best ROAS creatives.
Others will die between week 2 and week 3. This is just how selection on meta works and you can’t do anything about it.
Your goal is to do whatever it takes to find these 2-3 long term winners - again and again and again. This is why it takes time to get to high roas consistently. And this is also why starting with small budgets is a bad idea. The smaller the budget, the less creatives you will be able to test, the more time it will take you to reach a level where your meta roas becomes consistent & predictable.
The more variations and types of creatives you give to meta, the better it will understand your brand, and the higher chances you will have of finding your top evergreen creatives.
Do not trust meta attribution blindly. Use Shopify for total sales and meta only to track your spends.
For every 5,00,000 that you spend, 500-800 people will search your brand on Amazon. At a 5% (Amazon brand search CVRs are higher) conversion rate, you are looking at 25-40 sales on Amazon. Not being on Amazon can prove to be a mistake for just this reason alone, especially if you’re a high AOV brand.
Videos will take up 60% of your budgets.
The right way to creative testing for reels is by first uploading them as an organic post on your IG. you can use the trial reel feature or use a dummy page.
What works here for reach and engagement will work in your ads too.
Most brands do not do this because they are lazy. The ones that do this see a 10-15% lower CAC across the board.
Step 2 in creative testing is identifying a soft KPI for your ads performance. If you know what your best ads’ CPM is and your average A2C% (add-to-cart%) is, then all you need is to spend 5,000 on one creative before discarding it. This means that you do NOT need to spend a lot of money on a creative to understand if it will work or not.
Do not confuse creative testing with experiments. Creative testing is a BAU (business as usual) part of the process. 10-15% budgets should always go to creative testing. This is not optional.
But this 2nd part is entirely optional - and I strongly recommend it.
Your success in year 2 is directly proportionate to how many experiments you have conducted in year 1. Be sure to keep a 10% experimenting budget to try our various concepts, whitelisting, tech & landing pages.
During festive season, the brand which can spend the highest in BOFU campaigns wins. Small brands typically cannot spend more than 20% of the total budget in BOFU ever. However, large brands spend 3x more on TOFU, BAU - so that when come festive time, they can capitalise on these (often perceived as wasted) spends and retarget their existing & engaged customers aggressively.
As your brand matures, you will realise that festive performance gets better. In the beginning though, you will be seriously hit whenever other brands are doing their major sales.
It is important to make a list of the most relevant festivals and events for your brand and ensure that you are backwards planning for the same months in advance.
There are 2 types of changes you can make to a campaign:
Major changes are changes in ad budgets(more than 20%), any change in creative or delivery type.
Small changes are changes in captions, headlines, roas caps, bid caps etc.
You can make small changes as many times as you want, but big changes should only be done once per campaign. Every time a major change is made, the ad auction resets. This means that a top-performing ad may suddenly stop performing just because you made a major change.
This is why we recommend not increasing budgets by more than 20% at a time unless you’re running a limited-time offer or need to urgently scale.
Did you know that Indians spends 60% of their annual shopping budgets during Q4 of every year. Plan accordingly. This will either be your best time or worst time depending on what category you’re in. Typically giftable brands do very well during this phase.
Track roas for new users & repeat users separately to get a proper understanding of your numbers. Not only can this be tracked in analytics, you must create a segment for this on meta dashboard and optimise most metrics for new users. Let your retention & CRM take care of the repeats.
I am emphasizing on the importance of CAT again.
Your meta roas is directly proportionate to the consistent investments you make in content & tech. We see the highest returns when the founder owns the tech and not just outsources it to an agency.
Tech in 2025 is more than hygiene. It is a channel you must keep optimising 24*7, just like you expect your ads to get optimized.
As a founder, you don’t need to know or overthink the technicals of the meta dashboard. Just focus on deciding what your hero products are, what % of spending happens on which product and what stage of the funnel, and most importantly what kinds of creatives you’re running, and their ROI.
You should be speaking to your creative strategist either daily or at least twice a week. It’s okay if you’re checking in with the ads-execution team once a month to align on broad strategy.
The worst thing you can do for your time is try to micro-manage ad operations.
Apart from this, you should be consistently pushing the team to do more experiments and you should definitely micro-analyse the learnings of the creative testing campaigns.
I may have projected a somber narrative about meta ads and the unpredictable nature of this untamable beast, not all is dire.
While the business costs usually keep rising, we will often see tectonic cost reductions every few years.
-> Earlier there was no Shopify - so brands had to invest a lot in tech teams. Today, the cost of setting up a store is non-existent.
-> UGC videos were discovered a few years ago and the early days of UFC content saw CACs fall by as much as 50% for several brands.
-> AI is now being used widely which can lead to higher efficiency for content & design teams.
-> One-click checkouts with risk assessment came onto the scene 2 years ago giving relief to
-> First-party data could prove to be the next big thing in D2C that have the potential to reduce CACs by upto 60%.
As you can see, not all is gloomy. Every few years, through some technological or marketing breakthrough, brands find a way to reduce their cost of advertising - but the advantage only lasts for a few months and then diminishes as more brands jump on the bandwagon. This makes it absolutely essential for brands and their marketing partners to stay on top of the latest trends in marketing.
That's one of the biggest advantages of working with Porcellia. We're one of the few marketing teams out there that does not just rely on following & implementing new trends. We actively & intentionally keep experimenting so that we can discover new patterns and shape trends for the entire D2C ecosystem.
Case-in-point -> Here's how we cracked Reddit for D2C a few years ago
From D2C & Interested in staying on top of growth trends?
Everyone is talking about their wins,
but there is so much more you can learn from our failures.
Welcome to Finsweet's accessible modal component for Webflow Libraries. This modal uses Webflow Interactions to open and close. It is accessible through custom attributes and custom JavaScript added in the embed block of the component. If you're interested in how this is built, check out the Attributes documentation page for this modal component.
See docs