Andrei leads your Meta Ads operation: customer research, creative strategy, copy, static and video production, campaign deployment and ongoing buying. We manage the account and use its delivery evidence to decide what to make and test next.
ResearchCustomer language and approved product facts.
02
PrepareA concept, copy and production brief.
03
ReviewClaims, format, rights, costs and the next decision.
USD 5,000/month retainer + ad-spend fee. Ad budget and software/provider costs are separate.
From customer evidence to finished work.
Andrei leads the full Meta channel: ongoing media buying, customer research, creative strategy, copy, static/video production, deployment and optimization.
Full-service Meta · the work behind the account
Workstream
Work you receive
Operating focus
Customer and product research
A product-grounded creative direction, customer questions and concept briefs.
The product facts, audience problem, research sources and claims we can use.
Static and video production
Reviewed copy, visual directions, scripts and agreed finished formats.
Formats, rights, localization, production budget and review owner.
Testing and account management
Ongoing media buying, campaign deployment, testing and account optimization.
Budget allocation, creative delivery and the next useful testing decision.
Readout and next brief
An interpretation of covered delivery connected to the actual creative.
Whether to keep, revise or reject a concept; uncertain evidence stays uncertain.
Bring the brief and the buying context.
You bring product facts, access, brand assets and goals. We handle the research, production, media buying and ongoing optimization. Creative review is available when you want it.
Full channel ownership
The buying and the making stay together.
Creative needs delivery; media buying needs new creative and useful learning. Andrei owns that loop as one ongoing program. A creative-only request can be discussed separately.
The creative, ready to inspect.
Inspect the work itself. Open a piece at full size; captions distinguish complete page captures, historical openings and prepared examples. Displayed offers belong to the original artifact. Account results are not attributed to an individual design.
IL MAKIAGE · A restrained product-and-shade composition
A restrained product-and-shade composition. Historical AI-assisted portfolio creative. Depicted offers and claims belong to the original design, not a current offer or measured case result.
A balloon-led concept gives a fiber product a distinct visual language. AI-assisted creative; portfolio example, with no asset-level performance claim.
An expressive zipper-led product concept. Historical AI-assisted portfolio creative. Depicted offers and claims belong to the original design, not a current offer or measured case result.
A color-led shade comparison. Historical AI-assisted portfolio creative. Depicted offers and claims belong to the original design, not a current offer or measured case result.
A wide native-ad concept makes the product-format comparison visible. Historical AI-assisted creative; depicted formulation claims belong to the retained artwork, not an independently verified result or current offer.
A luxury flat-lay treatment. Historical AI-assisted portfolio creative. Depicted offers and claims belong to the original design, not a current offer or measured case result.
Selected retained production work, with each artifact’s original context. The measured Meta example follows with its own source and reporting window.
Follow a creative into the learning loop.
The retained brief, launch record and exact ad-level read connect one RESIST concept to a real operator decision. You can inspect the work and the limits of the observation together.
Ethos Car Care · 1–20 October 2026 · Actual agency operating work
The retained work, step by step
01Buyer context
Address a coating buyer skeptical of another graphene marketing claim.
02Concept and production
A dawn/afternoon hood scene makes the strategist’s heat-and-water-spot argument visible.
03Exact creative read
The original ad’s 1–12 October observation stays separate from its later scaling copy.
04Operator decision
On 20 October, the operator selected this creative for scaling and iteration. Later production or results are not inferred.
Creative1589, retrieved from the exact named Meta ad on 6 November 2026. AI-created dramatization, not a laboratory before/after. The displayed offer is historical.Open the complete image →
Ad spend
$222.19USD · original ad only
Attributed purchase value
$445.18USD · 7 attributed purchases
Recorded ROAS
2.00×Attributed value ÷ spend · rounded
1–12 October 2026 · America/Los_Angeles · Native Meta read 2026-11-06. Explicitly requested seven-day click, one-day engaged video and one-day view; impression-date reporting; unified attribution setting disabled. One website pixel purchase action, without summing purchase aliases. The current ad-set setting was captured separately. Daily historical attribution-setting history remains unknown.
Compare the retained historical read
The earlier local series retained USD 5.99 of unknown-bidding spend. Its money tuple matches the later native API read; these are separate observations, not two results to add together. Some October12 rows were fetched October18 before the final click window elapsed. Kept as historical source context; the separate November6 native API observation below confirms the same money tuple after the window.
The selected launch copy matches the dated receipt and the retained ad read. Its authored concept draft stays separate.
Trace the source, review and handoff
What the retained records establish
Record
What you can inspect
Grounded creative direction
The September 30 record retains five knowledge-source hashes, followed by the strategy, authored copy and image brief.
Actual selected launch copy
The canonical request matches its receipt and the retained ad read. It uses concept 1’s batch message with paragraph spacing normalized; concept 2’s different draft is preserved separately.
Operator decision
The October 20 read records selection for scaling and iteration. It does not establish causal lift or a later result.
Image identity
The request records an original PNG hash. The displayed image is the later Meta representation; those are distinct files.
Packaged code context: edition 2.0.0, released November 4, 2026, from source df2132cce260. This later packaging record does not identify the code checkout that generated the historical sample. Its execution-time revision and complete human-edit attribution remain unbound.
Observed attributed performance, not causal lift or audited profit. The selected launch-copy chain is retained; the exact execution-time source revision and complete human-edit attribution remain unbound. Historical references were not a randomized control. Batch and launch rules differ, and daily historical attribution settings remain unknown. The later packaged code edition is documented separately.
Source: Retained operational files and dated source records · Private source and exact ad identifiers excluded. Current image retrieval matches the named creative; original-launch image byte identity is unavailable.
Bloomday: Turn the research into a testable acquisition plan.
Research prepared for Bloomday connects buyer language, search demand and prospective offer economics. This was research prepared for the brand, not a client engagement or an achieved campaign result.
October 27, 2026 · research snapshot · Prepared research
YouTube comments
2,216Documented research corpus
Reddit records
414Separate research source; not buyers
Verified keywords
206Search research, not reached audience
Counts stated in the prepared research document. No client engagement, campaign execution or achieved result is claimed. Planning CAC is an assumption, not observed acquisition cost.
Compare a story-led car-care explanation with a short hands-free-light demonstration. These retained production cuts include captions and transcripts; their original offers stay with the artifact. No campaign result is attributed to either video.
Two products. Two ways to explain them.
Inspect the work itself. Open a piece at full size; captions distinguish complete page captures, historical openings and prepared examples. Displayed offers belong to the original artifact. Account results are not attributed to an individual design.
Ethos Car Care · Foam before the first wipe
A story-led Foam Cannon Pro V2 video connects the paint-care problem to a product demonstration. Production example; no performance result is attributed to this cut. Offer details belong to the original cut.
Those circles in your paint? You put them there on wash day. I wash trucks for a living, and every week someone pulls in with those fine little swirls you only see in the sun. That's grit. It sits on the paint, and the first thing that touches it, drags it around. A brush, a sponge, even your mitt. So nothing touches the paint until the foam's done its job. We cover the whole truck first. The foam loosens the dirt, so when the mitt goes on, there's less to drag. That's what we run here. The Ethos Foam Cannon Pro V2. It's rated up to 50% denser than a standard foam cannon, so it hangs on the paint instead of running off. If your pressure washer has a quarter inch quick connect, gas, or electric, it clicks right on. Right now, Ethos has it with a bottle of foam cannon soap for $74.99 instead of $84.98 separately. 30 days, money back. Click the link to see if the deal is still available.
Source: Tegra video production archive
SolderStick · Hands-free light under the hood
A short BeamSight demonstration recut shows the use case directly. Production example; no performance result is attributed to this cut. Offer details belong to the original cut.
Headlamps are just better? Nah. The light's built into the glasses at eye level, so no strap and no re-aiming my head.
BeamSight. Forty-nine ninety-nine.
Source: Tegra video production archive
Give the next batch a reason.
Customer language and product facts shape the concept. Reviewed production becomes a comparison; the covered readout gives the next brief a source.
Research → production → testing → learningIllustrative workflow
Customer problem → concept directions
An insight worth building around
A customer problem, the language behind it and distinct concept directions. See why one direction earns the brief and another is rejected.
The readout informs the next brief. The team owns execution; client creative review is available when requested.
Read the complete delivery sequence
An insight worth building around
A customer problem, the language behind it and distinct concept directions. See why one direction earns the brief and another is rejected.
Briefs and assets your team can review
Copy, visual direction, scripts and the agreed static/video outputs, checked for product accuracy, brand fit and rights. Production costs are agreed upfront.
A plan that answers a useful question
Connect the creative comparison to the audience, placement, budget and review point. We handle media buying, testing and production together.
A clear interpretation of the test
Understand what received delivery, how the concepts compared and whether there is enough evidence to decide. Keep uncertain results visible rather than declaring every test a winner.
A better brief for what comes next
A recommendation to keep, reject or revise the concept, with the reasoning your strategist and media buyer can use.
A comparison to agree before launch
Illustrative test plan
01
Compare two reviewed creative concepts
planned
For the same product and audience, which concept provides the stronger basis for the next creative batch?
Decision rule
Before launch, agree the concepts, audience, placement, budget, commercial measure and review point. Compare the covered delivery against that written rule. Recommend keeping, rejecting or revising a concept only when the evidence supports the decision; limited delivery stays uncertain.
Schedule
Started UnscheduledReview UnscheduledEnded Not supplied
Outcome
Pending; no outcome recorded.
This is an illustrative, unscheduled comparison. Full Meta management connects production, media buying and testing; no delivery or outcome is recorded in this illustration.
Source: Creative testing scope · Illustrative and unscheduled; the operating plan follows your account and commercial goals.
Give the next decision a reason.
How a test readout can guide the next batch. These are decision examples, not current account results.
Recommendation
What the readout needs to establish
Next creative decision
Keep
The covered evidence supports keeping the concept under the agreed decision rule.
Use the finding to inform the next brief.
Revise / learn more
Results are uncertain, delivery is limited or the concept needs a clearer comparison.
Improve the test or revise the concept before deciding.
Reject
The covered evidence supports rejecting the concept under the agreed decision rule.
Record the reasoning and change direction.
What stays attached to each creative decision
Readout records · our operating method
Record
Why it matters
Concept and creative version
Connect the interpretation to the actual brief and finished asset.
Observation context
Retain the window, attribution and unit before interpreting results.
Copies and delivery changes
Keep later versions and changed delivery separate from the original observation.
Decision and next brief
Record the reason for the next iteration and what still needs checking.
Your goals and access. Our team on the work.
Andrei Lunev leads the ongoing Meta Ads work. Your goals, account access, product facts and brand assets. We handle the work; creative review can be part of the workflow when you want it.
Typically live within 48 hours
Once account access and the required assets are ready.
Daily and hourly monitoring, with ongoing testing and optimization.
250–1,000 image and video ads per month
Production follows the account's opportunities and testing plan, with campaign and destination work connected.
Typical production reflects the opportunity and testing plan; the mix varies by account. Production volume is not a promise of a particular return.
For businesses ready to put more behind growth. Ecommerce is our strongest body of work; we also work with SaaS, apps and local businesses.
See how the fee adds up.
Google Ads and Meta Ads each have a $5,000/month retainer. When we manage both, the spend percentage applies once to their combined ad spend. Email/SMS is a separate $5,000/month flat fee.
Google / Meta · percentage on monthly ad spend
Combined managed ad spend
Ad-spend fee
Under $100,000
10%
$100,000 to under $250,000
7%
$250,000 to under $500,000
5%
$500,000+
4%
Bars: 0–10 percentage points. The monthly rate applies to the full managed Google/Meta spend, in addition to the channel retainers.
Your channels, one monthly example.
Illustrative
$5,0001 × $5,000 retainer+$10,5007% of $150,000=$15,500Service fee / month
What makes up the service fee
1 ad-channel retainer: $5,000
Ad-spend fee: $10,500
$15,500 total service fee · 1 managed channel / month
Each managed channel has its own $5,000 monthly retainer, invoiced when work begins. One ad-spend fee applies to the full actual combined managed Google and Meta spend, using that month's tier. It is invoiced at the beginning of the following month for the previous month. Email/SMS adds its own $5,000 retainer and no ad-spend percentage. Service prices include applicable tax.
The service total excludes your ad spend and separate software/provider costs.
From one example to work worth owning.
A useful example shows how we work. The next decision is whether you want ongoing execution, one bounded implementation or expert help with a specific question.
1
Start with a specific example
Bring a current campaign, creative, product page or customer journey and the decision behind it.
Is this an isolated correction or recurring work?
2
Choose the responsibility
Discuss what your team wants to keep and what you want Tegra to run across the selected channel.
Does ongoing managed execution fit?
3
Agree the first scope
Connect the initial work to a baseline, named owners, required inputs and a useful review window.
What will we review before deciding the next change?
What to align before work starts
The problem and the exact work in scope.
The responsible people, access and approval process.
The agreed fees, timing and acceptance or review measure.
A sample demonstrates the work. Your account data and agreed measurement determine how the commercial result is evaluated.
The account and creative, together.
Does this include media buying as well as creative?
Ongoing media buying, customer research, creative strategy, copy, static and video production, campaign deployment and optimization.
How much creative do you produce?
250–1,000 image and video ads per month. Production follows the account's opportunities and testing plan, with campaign and destination work connected. The mix follows the account, rather than a promise that every variant will perform.
Do we have to approve every creative?
No. You provide assets, product facts, brand rules, access and goals. We own ongoing execution. Creative review can be part of the workflow when you want it; your team does not need to manage production.
How do you decide what to make next?
We connect the actual creative to the delivery evidence, then keep, revise or change the concept. The customer problem, commercial measure and available evidence shape the next brief; a polished asset alone is not a winning result.
Can we discuss creative production for our own media team?
Yes. Tell us what your team runs and the production or testing support it needs. A creative-only request is discussed separately from the full managed-channel fee.
Start with the work you need.
Bring your current setup, goals and the work you want off your plate. In free discovery, we discuss fit, the channel scope and fees. If there is a fit, the written proposal confirms responsibilities, required access, production and review arrangements.
