Diana Nekrasova

More spend does not mean less work.

• In a post recounting a debate with young founders, they say a winning campaign can fall from 6.28 ROAS to 1.66, while cost per purchase can rise from EUR 5.92 to EUR 16.90 for a product with a EUR 35 average order value, in under a month if iteration stops.
• The post says global unicorn ads usually last 3-4 weeks; $2,000 budgets can get by with a few creatives, while $20,000-$50,000 budgets require a content factory.

A campaign that stops iterating can lose most of its profitability within weeks. The scaling question is whether the team can replace fatigued creative faster than efficiency decays.

Read original

Platform numbers diverge from CRM records: Facebook shows 100 conversions while the CRM records 65, and Google Ads claims 80 while analytics sees 52.

• Client-side pixels miss an estimated 20-40% of conversions because of ad blockers and consent rejections.
• Meta's January default shift to 7-day click and 1-day view means year-over-year comparisons need re-baselining.
• Their priority order: server-side tracking through CAPI alongside the pixel, hashed email on purchase, RFM segments from Klaviyo to Meta, then incrementality or geo-testing.

If analytics shows 100 conversions and the CRM shows 150 customers, Nekrasova says to fix the data gap before changing targeting or creative.

Read original

Platform-reported CAC and all-in CAC can differ by 20-40% once agency fees, creative production, influencer costs, and affiliates are included.

• They cite these thresholds: LTV:CAC below 1.5:1 is net negative; 2.0:1 is break-even but fragile; and 2.5:1 to 4:1 on a 12-month verified-cohort basis is healthy to scale from.
• Payback under six months is excellent; under twelve is workable.
• A 2026 study they cite found organic-dominant brands with 41% lower median CAC and 2.4x higher LTV:CAC than paid-dominant brands.
• Before scaling, they want a second-purchase rate of 25% or higher within 90 days and positive contribution margin per order.

A founder making budget decisions on platform-reported CAC may be looking at a number 20-40% smaller than the all-in cost of acquisition.

Read original

In one European e-commerce audit, they found Smart Sending suppressing emails and distorting analytics; email and SMS sending the same offer with no way to measure incremental lift; one flow producing zero revenue because of a logical conflict; and over-investment in acquisition.

• The fixes: stop unnecessary SMS spend, remove conflicting flows, and separate channels in the welcome flow.
• In a separate post, they argue most DTC brands have a post-purchase flow, but almost none have a good one.
• The sequence usually ends after shipping confirmation and resumes with the next promotion, even though post-purchase reaches customers when engagement is highest.

A flow can show zero revenue because of a logical conflict in its architecture, not because the message is weak.

Read original · Original 2

Examples include a Klaviyo welcome series on a $35K per month store typically adding $3-5K monthly, with first sales inside a week; modular ad creative built as 5 hooks x 3 visual concepts x 2 CTAs, tested at roughly $50-100 per day per test; and a content split around 70/30, with AI producing most output and humans handling brand voice, founder story, and trust-sensitive material.

• They distinguish verification from authority: an AI agent can check claims and fetch primary sources, but a human still decides whether to publish, name someone, or use a sensitive example.
• They cite a process failure in which an article went live without its fact-check step because it was published straight from the CMS.

The modular test structure—roughly $50-100 per day per test—shows how Nekrasova turns AI-assisted production into a repeatable testing system, not a one-off draft tool.

Read original

The paper, by Smith, Shulman and Kim, models an accuracy-parity policy in which users voluntarily disclose demographic identity.

• Under some conditions, the fairness policy can increase the firm's AI investment and benefit the dominant demographic group, while the smaller group it was designed to protect ends up worse off.
• Nekrasova notes the paper is paywalled and describes the post as a first reaction based on the abstract.
• Their argument: with Advantage+ and broad targeting, brands trust the algorithm to find the right customer, but for niche DTC brands the needed audience may be smaller, less represented in the data, or behave differently from the majority.

A niche DTC brand using broad Meta targeting may never realize that the algorithm does not fully understand its customers.

Read original

One post warns that DTC brands lose money in Q4 when inventory runs out while Meta campaigns keep spending; media buyers and operations teams should align inventory forecasting with ad spend before October.

• Another argues that the easy e-commerce era of AliExpress dropshipping and price-only competition is over because of rising tariffs, geopolitical shifts, and shrinking margins.
• They recommend diversifying sourcing into Turkey, Vietnam, India, Mexico, Colombia, and Peru; shifting toward premium or niche positioning; building a brand; improving logistics; expanding beyond the US; and tracking tariff shifts.
• A third post lists six e-commerce models: dropshipping, reselling, white-label or private-label, manufacturing, digital, and handcrafted.

Running Meta Ads into a stockout means paying for clicks and conversions, then absorbing cancellations, late shipping, refunds, and damaged trust.

Read original · Original 2 · Original 3

This is the fastest starting point.

Nekrasova says SciGrowth grew from a personal project into a full agency for Shopify brands. Its stated services are paid ads, CRM automation, and data-driven strategy, delivered by analysts, designers, media buyers, and CRM specialists.

For brands planning a campaign

Brands can treat SciGrowth as an integrated paid-media and retention partner, not a single-channel freelancer. Its Meta buying and Klaviyo capabilities make it a candidate for teams seeking support across both channels.

This post answers with numbers.

Nekrasova argues that scaling does not reduce work because creative fatigue sets in.

• They cite a campaign falling from 6.28 ROAS to 1.66 in under a month, cost per purchase rising from EUR 5.92 to EUR 16.90, and a EUR 35 average order value that wiped out the margin.
• Winning ads, they say, have a shelf life of roughly 3-4 weeks, making a content factory essential to scaling.

For brands planning a campaign

For a niche DTC brand past $20k/month, an audit should check creative output rate and margin at the actual AOV before increasing spend. The key question is whether the team can replace fatigued creative faster than efficiency decays.

This explains the measurement problem before targeting debates.

Nekrasova compares platform numbers with CRM data: Facebook shows 100 conversions while CRM records 65; Google Ads claims 80 while analytics sees 52.

• They estimate client-side pixels miss 20-40% of conversions because of ad blockers and consent rejections.
• Their recommendations include server-side tracking through CAPI, hashed email on purchase, RFM segments from Klaviyo to Meta, and incrementality or geo-testing.
• They also note that Meta changed default view attribution windows in January.

For brands planning a campaign

An audit should start by comparing platform-reported conversions with CRM records. If the gap is large, prioritize server-side tracking and Klaviyo-to-Meta segments over new targeting tests.

This post gives concrete thresholds.

Nekrasova says platform-reported CAC and all-in CAC can differ by 20-40% once agency fees, creative production, influencer costs, and affiliates are included.

• They classify LTV:CAC below 1.5:1 as net negative, 2.0:1 as fragile, and 2.5:1 to 4:1 on a 12-month verified-cohort basis as the healthy range.
• CAC payback under six months is excellent; under twelve is workable.
• They also cite a 2026 study finding organic-dominant brands with 41% lower median CAC and 2.4x higher LTV:CAC than paid-dominant brands.

For brands planning a campaign

Before scaling, require a verified cohort view, a second-purchase rate of 25%+ within 90 days, and positive contribution margin per order. This helps prevent media buyers from optimizing to platform ROAS while the business loses money.

This is the clearest example.

Nekrasova reads a Marketing Science paper on fairness in AI recommendation and connects it to paid media.

• With Advantage+ and broad targeting, brands trust the algorithm to find the right customer.
• They ask what happens when the needed audience is smaller, less represented in the data, or behaves differently from the majority.
• The paper is paywalled, so this is a first reaction based on its abstract.

For brands planning a campaign

For a niche DTC brand, an audit should test whether broad targeting reaches the customers the brand needs, not just the majority the algorithm understands best. The result could be a separate test for niche segments or a review of audience inputs, rather than an assumption that Advantage+ is always sufficient.

This gives practical numbers and a process warning.

Nekrasova says the approach matters more than the tools.

• A Klaviyo welcome series on a $35K/month store typically adds $3-5K monthly, with first sales inside a week.
• Modular ad creative uses 5 hooks x 3 visual concepts x 2 CTAs, tested at roughly $50-100/day per test.
• Content is split about 70/30 between AI and human work, with humans handling brand voice, founder story, and trust-sensitive material.
• They distinguish AI verification from human authority to publish, and describe an article that went live without its fact-check step because the CMS bypassed that check.

For brands planning a campaign

A lean DTC team can put this approach to work: retention flows first, low-cost modular creative tests second, and a documented human approval gate before publishing. The warning matters for brands adopting AI at scale.

A concrete example.

Nekrasova audited a Klaviyo account for a B2C e-commerce brand in Europe and found problems beyond email performance: duplicate triggers, Smart Sending suppressing emails and distorting analytics, email and SMS sending the same offer simultaneously, a logical conflict that left one flow generating zero revenue, and over-investment in acquisition. Fixes included stopping unnecessary SMS spend, removing conflicting flows, and separating channels in the welcome flow to measure incremental lift.

For brands planning a campaign

For a brand scaling past $20k/month, a retention audit should inspect flow logic and channel overlap, not just campaign performance. That can uncover wasted spend and lost LTV before acquisition budget increases.

This post identifies the gap.

Nekrasova says most DTC brands have a post-purchase flow, but almost none have a good one.

• A typical setup sends a shipping confirmation, perhaps an order summary, then stays silent until the next promotion.
• They argue post-purchase flows outperform other retention levers on cost per dollar recovered because they reach customers when engagement is highest, and point to a full 2026 playbook on sequence architecture, benchmarks, and priorities.

For brands planning a campaign

A DTC brand's campaign brief should include post-purchase as a measurable retention workstream, not just top-of-funnel ads. It can recover revenue at relatively low cost before the brand spends more on acquisition.

This explains the link to operations.

Nekrasova says many DTC brands lose thousands in Q4 when inventory runs out but campaigns keep spending.

• Running Meta Ads into a stockout means paying for clicks and conversions, then cancelling orders or shipping late and absorbing customer-service costs, refunds, and damaged trust.
• The fix is to align inventory forecasting with ad spend before October.

For brands planning a campaign

For a peak-season campaign, an audit should connect media buying to inventory data. A media buyer who coordinates with operations can reduce wasted spend and protect margin.

This is a simple example of their explanatory style.

Nekrasova lists six e-commerce business models: dropshipping, reselling, white-label or private-label, manufacturing, digital products, and handcrafted products. Each gets a one- or two-sentence definition, and the post ends by asking which models readers have worked with.

For brands planning a campaign

Useful for a brand or agency that needs shared vocabulary before a campaign, this post also shows Nekrasova's teaching approach: define the model, then connect it to operating choices.

This shows their strategic thinking beyond ad accounts.

Nekrasova argues that the easy e-commerce era of AliExpress dropshipping and competing only on price is over, as tariffs rise and margins shrink. They recommend diversifying sourcing beyond China into Turkey, Vietnam, India, and Latin America; shifting from mass-market to premium or niche products; building a brand; optimizing logistics and cutting waste; expanding beyond the US into the Middle East, Latin America, and Southeast Asia; and tracking tariff and policy shifts.

For brands planning a campaign

For brands exposed to tariffs, a campaign can shift from price-led acquisition to premium positioning and new geographies. Nekrasova connects media strategy to supply-chain and margin realities when rising costs change the offer.

In one post, they describe debating young founders who believed more budget meant less work.

• They counter that, if iteration stops, a winning ad’s ROAS can fall from 6.28 to 1.66 and cost per purchase can triple from €5.92 to €16.90 for a €35 AOV product in under a month.
• The post says global “unicorn” ads usually have a 3–4 week shelf life; $2,000 budgets can get by with a few creatives, while $20,000–$50,000 budgets require a content factory.

A content-factory approach to creative iteration and scaling.

For brands planning a campaign

For a DTC paid-media audit, test creative iteration cadence and production capacity before increasing spend. Track marginal ROAS and cost per purchase as budgets scale; don’t assume efficiency will remain linear.

Examples: Source 1

The post notes that, under some conditions, a fairness policy can increase the firm’s AI investment and benefit the dominant demographic group, while the smaller group the policy was designed to protect ends up worse off.

• They connect this to paid media: with Advantage+ and broad targeting, brands trust the algorithm to find their audience.
• That raises questions about what happens when a niche DTC brand’s audience is smaller, less represented in the data, or behaves differently from the majority.

Nekrasova states the paper is paywalled and the post is based on the published abstract; it is a first reaction, not a full review.

For brands planning a campaign

For a niche DTC account, check whether broad-targeting results mask underperformance among smaller or less-represented audience segments. Run segment-level or geo-incrementality checks before treating broad targeting as universally safe.

Examples: Source 1 · Source 2

The post describes gaps between platforms and CRM data: Facebook shows 100 conversions while the CRM records 65, and Google Ads claims 80 sales while analytics sees 52.

• Client-side pixels are estimated to miss 20–40% of conversions from ad blockers and consent rejections alone.
• Meta’s January default shift to 7-day click and 1-day view means year-over-year comparisons need re-baselining.
• They prioritize server-side tracking through CAPI alongside the pixel, hashed email on purchase, RFM segments synced from Klaviyo to Meta, and incrementality or geo-testing once that foundation is solid.

For brands planning a campaign

Start an audit by reconciling measurement: compare platform-reported numbers with CRM data, check CAPI coverage and hashed-email match rates, and account for attribution-window changes before optimizing targeting or creative.

Examples: Source 1

The post says platform-reported CAC and all-in CAC can differ by 20–40% once agency fees, creative production, influencer costs, and affiliates are included.

• Benchmarks cited: LTV:CAC below 1.5:1 is net negative territory; 2.0:1 is break-even but fragile; 2.5:1 to 4:1 on a 12-month verified-cohort basis is the healthy range to scale from; CAC payback under 6 months is excellent, and under 12 is workable.
• They also cite a 2026 study finding organic-dominant brands running 41% lower median CAC and 2.4x higher LTV:CAC than paid-dominant brands in the same sample.
• Before scaling, they want a second-purchase rate of 25%+ within 90 days and positive contribution margin.

For brands planning a campaign

Before recommending a paid-budget increase, require all-in CAC, verified cohort LTV, payback period, second-purchase rate, and contribution margin—not just ROAS.

Examples: Source 1

Findings included Smart Sending suppressing emails and distorting analytics; email and SMS sending the same offer with no way to measure incremental lift; one flow producing zero revenue because of a logical conflict in the architecture; and over-investment in acquisition relative to retention.

• In a separate post, they say most DTC brands have a post-purchase flow but almost none have a good one: usually shipping confirmation, maybe an order summary, then silence until the next promotion.
• They describe post-purchase as consistently outperforming other retention levers on cost-per-dollar-recovered.
• Another post calls Klaviyo flows the fastest ROI path, saying a welcome series on a $35K/month store typically adds $3–5K monthly, with first sales inside a week.

Klaviyo architecture audit: trigger logic, flow conflicts, channel separation, and segmentation.

For brands planning a campaign

For a DTC brief, audit flow logic, channel overlap, SMS spend, welcome and post-purchase sequences, and incremental lift—not just open rates or list size.

Examples: Source 1 · Source 2 · Source 3

The post describes modular ad creative built as 5 hooks x 3 visual concepts x 2 CTAs, generated quickly and tested cheaply at roughly $50–100/day per test.

• It recommends a 70/30 content split, with AI handling most output and humans handling brand voice, founder story, and trust-sensitive material.
• They distinguish verification from authority: an AI agent can adversarially check every claim in a draft and fetch primary sources, but a human still needs to decide whether to publish, name someone, or use a sensitive example.
• The post cites a process failure in which an article went live without its fact-check step because it was published straight from the CMS.

Modular creative testing with AI-assisted output and human approval for sensitive or authoritative decisions.

For brands planning a campaign

When evaluating AI capability, ask for the documented production process and human review gates, not a tool list. Test whether creative modularity and claim-checking are built into production.

Examples: Source 1 · Source 2 · Source 3

The action plan includes rethinking supply chains by exploring Turkey, Vietnam, India, Mexico, Colombia, and Peru; shifting from mass-market to premium or niche positioning with higher quality, sustainable or ethical branding, and specific pain points; building a brand rather than a store; optimizing logistics to cut waste; expanding beyond the US; and staying ahead of tariffs and disruptions.

• A separate post lists six e-commerce business models: dropshipping, reselling, white labeling or private labeling, manufacturing, digital products, and handcrafted products.
• Another warns that DTC brands lose money in Q4 when inventory runs out while Meta campaigns keep spending, and argues that media buyers and ops teams must align inventory forecasting with ad spend before October.

For brands planning a campaign

For e-commerce campaign planning, assess operational readiness: margin pressure, sourcing diversification, inventory alignment with ads, and whether the brand can sustain paid spend through stockouts or tariff changes.

Examples: Source 1 · Source 2 · Source 3

Its stated services are paid ads, CRM automation, and data-driven strategy for Shopify brands, with campaign conversion, Klaviyo flows, and clarity for founders as core goals.

• They identify as SciGrowth’s founder with 7+ years in global e-commerce.
• Posts offer audits of Klaviyo setups and link to full articles on scaling, attribution, and AI workflows.

For brands planning a campaign

The profile points to an agency-service and audit orientation, useful for briefs combining paid media, CRM, and measurement. Review linked articles and audit outputs rather than expecting individual creator-style sponsorship formats.

Examples: Source 1 · Source 2 · Source 3 · Source 4 · Source 5

As founder, they describe its analysts, designers, media buyers and CRM specialists, refer readers to scigrowth.com, and cite more than seven years in global e-commerce as their background.

Original disclosure

What started as my personal project has grown into something much bigger. Today we officially launch SciGrowth — a marketing agency helping Shopify brands grow through smart paid ads, CRM automation, and data-driven strategy.

View example

They say they spend most of their week in Klaviyo dashboards, call Klaviyo flows the fastest ROI path for lean DTC teams, and describe auditing the full retention architecture of a Klaviyo account — trigger logic, flow overlaps, attribution, SMS spend and segmentation — for an unnamed European e-commerce brand.

View example

Meta's targeting and measurement changes recur in their posts: Advantage+ and broad targeting, the January shift to 7-day click and 1-day view attribution defaults, and how niche DTC brands fare when an algorithm decides the audience. They also write about running Meta campaigns into stockouts.

View example

Aslam describes running Meta Ads for a fashion e-commerce brand as testing, iterating, and knowing when to scale or pull back. Nekrasova makes the same case with audit numbers, framing scaling as an engineering process that accounts for creative fatigue.

Aslam draws on a media buyer’s career and client relationships, including leaving a salaried role to build a freelance practice, giving readers the personal and business sides of buying.

• Nekrasova writes from the agency level, linking buying discipline to attribution rebuilds, Klaviyo flow architecture, and CAC and LTV thresholds.
• Career transitions are outside their coverage.

Read original · Original 2 · Original 3

Nekrasova warns that rising tariffs and shrinking margins end the era of buying from China on price alone. Vega argues that China+1 does not mean China-1, and that moving final assembly does not move the underlying component and tooling supply chain.

Vega focuses on sourcing and procurement, writing for people who map suppliers, tooling, and single-source dependencies.

• They make the case for nearshoring to Mexico on lead time and capability, rather than labor cost.
• Nekrasova ties sourcing to the marketing budget: whether a brand can keep paying for clicks through stockouts and tariff shifts, and how media buying should align with inventory planning.

Read original · Original 2 · Original 3 · Original 4

Hoover-Salomon’s post on the proposed tariffs aimed at low-cost platforms covers the same territory as Nekrasova’s argument that the older model of cheap goods from China is closing.

Hoover-Salomon writes from the freight and logistics perspective, covering shipping patterns, customs compliance, trade routes, and the shift toward zero-emission trucking to show how demand moves. Nekrasova writes from the brand and paid-media side, covering sourcing diversification, premium positioning, expansion into new regions, and the inventory-versus-ad-spend coordination that prevents Q4 losses.

Read original · Original 2 · Original 3 · Original 4

Diana Nekrasova's posts cover China+1 sourcing strategy, tier-one supplier mapping and nearshoring to Mexico.

• They argue that sourcing decisions should withstand changes in currency, tariffs, freight and labour.
• For a brand whose growth is constrained by supply or tariff-driven margin pressure, this expertise is more directly relevant.

The posts do not discuss paid social, creative testing, retention flows or measurement work. A campaign team seeking a media-buying and attribution audit would need a paid-media and lifecycle operator alongside this sourcing advice.

Examples: Source 1 · Source 2

Diana Nekrasova's posts address the logistics consequences of tariffs on platforms like AliExpress and Shein, and the case for zero-emission trucks in freight. For a brand rethinking where and how products ship, this is a different starting point from an ad account audit.

The posts address the movement of goods, not acquisition and retention. A campaign's media-buying and lifecycle work would need a separate specialist.

Examples: Source 1 · Source 2

Diana Nekrasova's posts describe testing, iterating, reading the data and choosing when to scale or pull back.

• They cite five-plus years of Meta Ads, fashion e-commerce and ROAS optimisation, including a recent in-house brand role.
• That fits a campaign seeking day-to-day account operation rather than a wider systems rebuild.

Diana Nekrasova's posts cover more of the surrounding system than Hamza Aslam's two posts, which are personal reflections that show media-buying experience but no retention, CRM or measurement architecture, and no team or agency scale. For attribution reconciliation, Klaviyo lifecycle flows, all-in CAC and payback thresholds, or a creative-factory cadence for scaling, ask for examples of the creator's work in those areas; a campaign may need a second specialist.

Examples: Source 1 · Source 2 · Source 3 · Source 4 · Source 5 · Source 6 · Source 7 · Source 8