BlueAds × Naturvia
Meta Ads: the numbers and the decisions
Naturvia started working with BlueAds in December 2025, from scratch, with no account history. The initial objective was a ROAS of 4 and increasing volume for as long as the business remained profitable.
Between 1 August and 12 October 2026, Meta Ads campaigns recorded 23,281 purchases with a total attributed value of RON 3,332,205.10. Advertising spend was RON 645,774.61, with an average cost per purchase of RON 27.74.
Behind these figures was a continuous testing process. We worked on creative, audiences and campaign structure, as well as the pages people reached after clicking. For the strongest-performing products, we tested new angles through combinations of different ads and landing pages, aligning the message, offer and website experience.
This study explains how we built the strategy and how results developed. The partnership began in December 2025; the figures below cover the 73 days from August to October 2026.
01 / CONTEXT
From an account with no history to sustained purchase volume
Naturvia is a dietary supplement brand selling online in Romania. When the partnership began in December 2025, we started with no account history. Testing was needed to identify the combinations of products, messages and audiences that could support sales.
The initial ROAS target of 4 was our benchmark for advertising efficiency. The business objective was to grow volume as much as possible while maintaining profitability. As the account accumulated results, scaling decisions could draw on campaign history and stability.
The challenge at high volumes
To grow, we needed new purchasing opportunities: new people to reach, messages relevant to them and a consistent experience after the click. Budget increases needed the support of this testing process.
In August and September, each month exceeded RON 1.3 million in purchase value attributed to Meta Ads. We aimed to maintain this volume and continue developing the account while keeping advertising efficiency above the initial benchmark.
02 / THE BLUEADS STRATEGY
A structure that supports testing and scaling in parallel
ABO and CBO, alongside different optimisation objectives
We used both ABO campaigns, with budgets at ad set level, and CBO campaigns, with budgets at campaign level. The account ran campaigns optimised to maximise the number of conversions and campaigns optimised to maximise conversion value.
These objectives pursue different outcomes: the first focuses on the number of conversions, while the second focuses on their value. We included both in the strategy to test purchase volume and value, assessing results against the account's objectives.
Fresh creative introduced consistently
Creative was central to most campaigns. We tested a large volume of material, particularly video, supported by banners and catalogue ads. Testing continued while active campaigns were maintained and scaled.
We explored different communication angles for the same product. An angle is the perspective from which we present a product and the reason it may be relevant to someone. The difference may lie in the usage context, the question we start with or the argument we develop.
This diversification helped us reach new audiences and expand the audience for existing products. New material remained part of account development after we had identified products that performed well.
Broad audiences, different targeting and remarketing
We combined broad audiences, without narrow targeting, with tests across different audiences and targeting options. We also included a remarketing campaign in the account structure.
For best-selling products, we also tested campaigns that excluded existing customers. These tests aimed to extend promotion to people who had not already bought from the brand.
03 / BEYOND CREATIVE
We also tested new angles through different landing pages
For the strongest-performing products, we extended testing to the entire buying journey: from the ad that attracts attention to the page that explains the offer.
A product can be relevant to people starting from different questions or motivations. If we change the perspective in the ad, the destination page needs to continue that explanation. Otherwise, visitors have to make the connection themselves between what attracted them and the offer they find on the website.
That is why we tested different landing pages alongside different creative for our strongest products. We focused on alignment between the communication angle, product presentation and offer. The page became part of the test, together with the ad and the audience receiving it.
What alignment between an ad and a page means
A consistent experience continues the idea that generated the click. Visitors should quickly recognise the subject of the ad, understand the offer and find the information they need to decide. The order of information and the emphasis placed on particular explanations can matter at this stage.
For Naturvia, page tests supported this connection. We could explore new approaches for proven products, expand the audiences we addressed and provide a relevant continuation of the ad's message.
- 01
Communication angle
Why the product is relevant.
- 02
Creative
The message that attracts attention and generates interest.
- 03
Landing page and offer
The explanation that continues the message after the click.
- 04
Measured result
Purchases, their value and the cost of acquiring them.
How this approach contributes to greater volume
In our view, testing combinations of creative and pages helped us find better ways to present the offer to new audiences. When the message is relevant and the website experience supports the buying decision, there is a stronger basis for conversion.
Measured interactions and conversions also provide signals for delivery optimisation. This is how we helped the algorithm identify purchasing opportunities: by testing consistent experiences and using the results they produced.
Account results reflect the combined effect of the strategy: pages, creative, audiences, budgets and campaigns. Aggregate reporting does not isolate each landing page's contribution; that would require a separate analysis of the tests.
04 / A BRIEF LOOK AT META'S ALGORITHM
How Meta decides who sees an ad
With the introduction of Andromeda, Meta improved how it initially selects candidate ads for each person. The system uses more advanced AI models to identify relevant ads from a large pool of options and personalise this selection more effectively.[1]
Meta uses machine learning models to select and rank relevant ads for each person. The system estimates likely outcomes using available signals about ads, context and behaviour. Selection involves several stages: identifying candidate ads, then evaluating them in more detail.[1]
The objective defines the outcome we pursue. For Naturvia, we used both maximising the number of conversions and maximising their value.
The system estimates relevance for each person. Meta describes models that use interactions and conversion events, including their order and timing, to interpret behaviour and make predictions.[2]
Measured results support optimisation. Website events shared through available measurement tools, such as Meta Pixel or Conversions API, can be used for reporting and optimisation.[3]
Testing provides new options to evaluate. In our strategy, different creative and landing pages aimed to make the offer relevant to more purchasing contexts.
A new landing page does not automatically communicate a precise audience to the algorithm. Its role is to support the experience and conversion after the click. The connection to optimisation comes through measured results and signals; performance depends on the entire journey. This explains the mechanism without assuming that every new variation will improve ROAS.
Technical sources
05 / CONTROLLED SCALING
Gradual budget increases on campaigns with a track record
We scaled in two ways: by launching new campaigns and by increasing budgets on active campaigns. Increases were controlled and gradual, applied when campaigns were stable, had a track record and met the parameters we were monitoring.
At the same time, we continued introducing fresh creative and testing angles, audiences and pages. This work supported account expansion while campaigns with proven results continued running.
Decisions focused on the relationship between budget, purchase volume and efficiency. A campaign with a track record gave us a stronger basis for increasing investment. New tests opened up other development opportunities for existing products.
06 / DOCUMENTED RESULTS
Over RON 3.3 million in sales attributed to Meta Ads
Over the 73 days analysed, each leu invested in advertising was associated with approximately RON 5.16 in purchase value reported by Meta. Overall ROAS exceeded the initial target of 4.
| Period | Spend (RON) | Attributed sales (RON) | Purchases | ROAS | CPA (RON) |
|---|---|---|---|---|---|
| August | 262,919.87 | 1,319,004.29 | 9,361 | 5.02 | 28.09 |
| September | 261,616.60 | 1,322,812.38 | 9,140 | 5.06 | 28.62 |
| 1–12 October | 121,238.14 | 690,388.43 | 4,780 | 5.69 | 25.36 |
| TOTAL | 645,774.61 | 3,332,205.10 | 23,281 | 5.16 | 27.74 |
CPA = cost per purchase. Overall ROAS and CPA are calculated from cumulative values, not the simple average of monthly metrics. October is a partial period.
01 / TOTAL PERIOD · 1 August–12 October 2026. Original Meta Ads Manager screenshot. Click to enlarge.August · over RON 1.3 million and a ROAS of 5.02
August recorded 9,361 purchases and attributed sales of RON 1,319,004.29, with spend of RON 262,919.87. Average cost per purchase was RON 28.09, while ROAS of 5.02 was above the initial target.
02 / AUGUST · Results for 1–31 August 2026.September · sustained sales volume, with a ROAS of 5.06
In September, purchase value reached RON 1,322,812.38, with 9,140 purchases and spend of RON 261,616.60. Attributed revenue was close to August's level on a slightly lower budget. ROAS rose to 5.06, while CPA was RON 28.62.
03 / SEPTEMBER · Results for 1–30 September 2026.1–12 October · ROAS of 5.69 and CPA of RON 25.36
During the reported October period, campaigns recorded 4,780 purchases with a value of RON 690,388.43 and spend of RON 121,238.14. ROAS rose to 5.69, while cost per purchase fell to RON 25.36.
04 / OCTOBER · Results for 1–12 October 2026, up to the time of the screenshot.07 / VOLUME AND EFFICIENCY
A higher sales pace alongside improving ROAS
Since October includes only 12 days, comparison with September needs to account for the different period lengths. Daily averages describe the recorded pace and are not a forecast for the rest of the month.
1–12 October compared with September 2026
| Metric | September · 30 days | 1–12 October · 12 days | Change |
|---|---|---|---|
| Average spend / day | RON 8,720.55 | RON 10,103.18 | +15.9% |
| Attributed sales / day | RON 44,093.75 | RON 57,532.37 | +30.5% |
| Average purchases / day | 304.7 | 398.3 | +30.7% |
| ROAS | 5.06 | 5.69 | +12.5% |
| CPA | RON 28.62 | RON 25.36 | −11.4% |
In the reported October period, average daily spend was approximately 15.9% higher, while average daily purchase value rose by 30.5%. Average daily purchases also increased, alongside an improvement in ROAS.
These results support the direction we pursued: expanding volume with better advertising efficiency. They reflect the account strategy as a whole. They are not an isolated measurement of landing page impact or a comparison between controlled groups.
Averages are calculated from reported totals divided by the number of calendar days. Changes in ROAS and CPA use the displayed metrics, rounded to two decimal places.
08 / WHAT WE ARE TESTING NEXT
Expanding audiences and preparing new products
Testing continues beyond the period analysed. We are still working on new audiences and expanding the audience for current products, alongside testing new products for launch.
More potential for existing products
For current products, we continue exploring new audiences and angles. A product that already performs well offers a starting point for other approaches, but each new direction needs to be assessed on its own results.
Creative and landing page testing remains relevant here. A message that works for one group of buyers may need a different explanation for a new audience. Maintaining continuity between the ad and the offer helps us build these tests around a consistent journey.
New products for launch
In parallel, we are testing new products for launch. The aim is to identify further opportunities to develop the portfolio and sales volume. Experience accumulated in the account provides a foundation, while each new product needs validation of its own offer, messages and audience.
The results of these tests and future launches are not included in this study's figures. They represent the next stage of work in the Naturvia × BlueAds partnership.
09 / CONCLUSIONS
Scaling is built across the entire buying journey
The Naturvia experience shows why testing should cover both the ad and the destination page. Creative introduces the message, the landing page develops it, and measured results help us evaluate the combination.
For the strongest-performing products, we used this approach to explore new angles and audiences. We supported the process with a diversified campaign structure and gradual budget increases on stable campaigns with a track record.
In 73 days, Naturvia recorded over RON 3.3 million in sales attributed to Meta Ads, 23,281 purchases and a ROAS of 5.16. During the reported October period, average daily volume grew alongside ROAS, while cost per purchase fell.
We continue testing new audiences, expanding the audience for existing products and preparing growth opportunities through new products.
