Measurement Layer

About Measurement Layer

Marketing measurement, unpacked. Kostya Vashkevich sits down with practitioners — heads of marketing, agency operators, tracking engineers — and unpacks one specific piece of measurement infrastructure per episode. How tracking stacks actually work. What attribution gets right and where it lies. How operators build measurement that holds up under scrutiny. Operator-to-operator. 15-30 minutes. No fluff. Host: Kostya Vashkevich, Head of Marketing and former media buyer. More than 10 years working on one problem: making marketing measurement actually work. New episode every month on YouTube, Spotify, and Apple Podcasts.

Publisher

Kanstantsin Vashkevich

Category

business · technology

Language

en

Explicit

No

First episode

30 Jun 2026

Latest episode

13 Sept 2026

Latest episodes

6 episodes in the feed.

1. 

   13 Sept 2026

   Why GA4 never matches your ad platform · Evelyn Välja

   Your numbers were never supposed to match. Evelyn Välja audits ecommerce tracking, and she says fixing the tags is the easy half. Evelyn spent twenty years in data governance at Swedbank, then five years running two restaurants in Finnish Lapland, and now audits measurement setups for small ecommerce teams. This episode covers what actually breaks in those setups: duplicated purchase events, consent mode that looks compliant but is not wired to anything, and GA4 revenue that never lines up with Meta or Google Ads. It is also about the part nobody fixes, which is what happens to a team after it stops believing its own reports. What we cover: - Why a working cookie banner tells you nothing about whether consent is applied to your tags - How one order ends up counted twice, and why that can run for months without anyone noticing - Why GA4, Meta and Google Ads were never built to report the same number, and how to explain that to an owner who expects a match - The sign during an audit that a team has already stopped trusting its data - What consent mode advanced actually recovers, and what it does not - The one habit from bank scale data governance a five person team should copy: agree the definitions before you argue about the numbers - The twenty minute check that finds most of it: follow one real purchase through every system you already use Chapters: (00:00) The numbers nobody trusts (02:54) The banner that was not connected (06:11) One order counted twice (07:22) Why GA4 never matches your ads (10:03) When a team stops trusting data (13:47) What consent mode really recovers (16:17) Agree definitions before the numbers (19:55) Follow one purchase next week Find Evelyn Välja on LinkedIn: https://www.linkedin.com/in/evelynvalja/ (https://www.linkedin.com/in/evelynvalja/) Digimeetrik: https://digimeetrik.ee/en/ (https://digimeetrik.ee/en/) Watch the full episode on YouTube: https://www.youtube.com/watch?v=XrlK8NaS-Mw (https://www.youtube.com/watch?v=XrlK8NaS-Mw) Hosted by Kostya Vashkevich. Work with me on growth marketing and measurement for B2B SaaS, or apply to be a guest: https://www.linkedin.com/in/konstantin-vashkevich/ (https://www.linkedin.com/in/konstantin-vashkevich/)

2. 

   17 Aug 2026

   Marketing measurement at scale · Roman Petrochenkov · EP05

   At CarWow's scale, tracking is the easy part. The hard problem is definitions. Roman Petrochenkov, Head of Analytics for marketing and media, on what measurement really takes at scale. This episode is about what marketing measurement demands once an organization is big enough to run a dedicated analytics function. Roman leads analytics for marketing and media at CarWow, a marketplace with millions of visitors and several verticals. We get into where the gap between what the business needs and what the data can deliver shows up first, why definitions are harder than tracking, how to align stakeholders without forcing agreement, and where server-side tracking, attribution, incrementality, MMM, and AI fit at that scale. What we cover: - Why tracking is a solvable technical exercise you can outsource, and definitions are the two-year fight you cannot - How a five-person analytics team plus an agency resource splits daily decisions from deep six-month projects - The server-side migration that lifted signal quality 10 percent, and the cross-device story behind a 120 percent YouTube conversion lift - Attribution, incrementality, and MMM as three separate questions, and the budget scale each one earns - Why AI gives a confident wrong answer on CAC without a semantic layer and a data warehouse underneath it Chapters: (00:00) Both sides of the measurement gap (02:53) Tracking is easy, definitions are hard (05:09) Aligning marketing, finance, and legal (11:29) Building the analytics team (16:26) Server-side first, 10% uplift (22:07) Attribution, incrementality, and MMM (27:20) AI and the semantic layer (35:27) One change to make next week Find Roman Petrochenkov on LinkedIn: https://www.linkedin.com/in/rpetrochenkov/ (https://www.linkedin.com/in/rpetrochenkov/) Watch the full episode on YouTube: https://youtu.be/dbJc3kqcHfU (https://youtu.be/dbJc3kqcHfU) Work with me on growth marketing and measurement for B2B SaaS, or apply to be a guest: https://www.measurement-layer.com/ (https://www.measurement-layer.com/) Hosted by Kostya Vashkevich. https://www.linkedin.com/in/konstantin-vashkevich/ (https://www.linkedin.com/in/konstantin-vashkevich/)

3. 

   6 Aug 2026

   Platform attribution is a weak signal · Alex Ignatenko, Fractional Head of Marketing Analytics

   Alex Ignatenko is a fractional Head of Marketing Analytics working with ecommerce, online education, and iGaming teams. His argument: the hardest part of attribution in 2026 is not the data. Every client arrives with the same symptom. GA4 says one number, Meta says another, the CRM says a third, and nobody can defend the gap. We spend the episode walking down to where that actually starts. It starts at taxonomy, not at the ad platform. You launch with fifteen clean events, someone adds one extra dimension for one specific case, and a year later there are three hundred events nobody in the building can explain. Fix that and you still cannot trust in-platform numbers. Meta will always attribute to Meta. Alex calls that a feature of a public company rather than a bug: fine for optimising campaigns, not for deciding budget. So you build your own layer, and that is where the real ceiling shows up. It is cognitive, not technical. Humans read two dimensions well. CPA on day one, CAC at 30, and LTV at 90 across every channel at once is not a layout problem, and no dashboard solves it. Which is why every dashboard is a model of reality, and reality is under no obligation to agree with it. When customers tell you something your model contradicts, believe the customers. We also get into what a data agent needs before you let it near attribution logic: a clean dataset, explicit attribution rules, a semantic layer, definitions, and verified queries. An agent given full freedom optimises to please you. You do not need to be pleased, you need a decision. Chapters: (00:00) Intro (00:32) Plenty of data, no agreement on what it means (03:24) Where trust breaks: tracking and taxonomy (06:42) Why platform attribution is a weak signal (09:34) Not a bug, a feature: how Meta's incentives shape your numbers (11:04) Technical expertise as the real bottleneck (16:37) The ceiling is cognitive, not technical (19:53) When a team stops reading the dashboard and starts guessing (22:18) Building attribution a data agent can actually use (26:46) Which queries earn verified status (29:29) When the data argues against the decision (30:47) Every dashboard is a model of reality (32:22) Takeaway: there is no magic pill (33:59) Where to find Alex Mentioned at 14:25: Simo Ahava's free material on how browsers actually handle tracking. Find Alex Ignatenko on LinkedIn: https://www.linkedin.com/in/aleksandr-ignatenko/ (https://www.linkedin.com/in/aleksandr-ignatenko/) Watch the video version on YouTube: https://www.youtube.com/channel/UCtiQFCikwcePibHHgf5Bodg (https://www.youtube.com/channel/UCtiQFCikwcePibHHgf5Bodg) Be my next guest, or work with me: https://www.measurement-layer.com/ (https://www.measurement-layer.com/) Hosted by Kostya Vashkevich. https://www.linkedin.com/in/konstantin-vashkevich/ (https://www.linkedin.com/in/konstantin-vashkevich/)

4. 

   25 Jul 2026

   How to measure organic and GEO when there's no click | Vladan Ubiparip, dramaticSEO | EP03

   "90% of GEO is just good SEO." — Vladan Ubiparip, founder of dramaticSEO. Vladan has been in SEO for 10 years, working exclusively with B2B SaaS companies above $1M ARR. In this episode, he drops several hot takes that most SEO practitioners won't say publicly — including why traffic is still a vanity metric, why most GEO measurement tools are irrelevant, and why 90% of GEO is just good SEO done properly. In this episode, Vladan explains: - Why traffic is a vanity metric — and what to measure instead - The most common organic attribution mistake B2B SaaS teams make right now - How to run a quick organic measurement diagnostic this week with no attribution setup - What GEO actually means for measurement — and why there's no click to track - Why AI visibility prompt tracking tools are "the most irrelevant metrics you can track" - The one tool worth paying attention to for AI brand visibility (Ahrefs brand monitor) - Why 90% of GEO optimization is just good SEO — the SEO Summit Sibenik story - The one event every SaaS team should set up for organic measurement: "SEO/GEO qualified lead" - Who inside a SaaS company should own organic measurement — and why it follows CAC ownership Timestamps: (00:00) Intro (01:04) What dramaticSEO does and who they work with (02:37) The state of organic tracking when companies pass $1M ARR (05:29) First thing Vladan checks in a new client's measurement setup (08:09) The most common attribution mistake in organic measurement (09:35) How internet consumption is changing and what that means for your metrics (11:42) Quick diagnostic: what to run if you have no attribution setup (14:40) Self-attribution surveys — why every SaaS needs one immediately (16:29) GEO vs SEO — what's actually different in measurement (18:55) The missing click: why GEO is harder to track than SEO (20:39) GEO measurement tools — what works and what's irrelevant (21:51) Ahrefs brand monitor — the only AI visibility tool Vladan recommends (26:01) The SEO Summit Sibenik story: 60-year-old SEO wins at GEO by doing basics (27:09) Where organic measurement breaks in the marketing stack (29:23) Who should own organic measurement inside a SaaS company (32:43) One change to make next week: create an "SEO/GEO qualified lead" event (36:40) Why organic should get first-touch credit — not last-touch Find Vladan: https://www.linkedin.com/in/vladanx/ (https://www.linkedin.com/in/vladanx/) dramaticSEO: https://dramaticseo.com/ (https://dramaticseo.com/) Measurement Layer is a podcast where I sit down with practitioners and unpack how marketing measurement actually gets built. New episode every month on YouTube, Spotify, and Apple Podcasts. Host: Konstantin Vashkevich. Ten years working on one problem: making marketing measurement actually work. Currently Head of Marketing at RedTrack.io (http://RedTrack.io). Find me: https://www.linkedin.com/in/konstantin-vashkevich/ (https://www.linkedin.com/in/konstantin-vashkevich/)

5. 

   11 Jul 2026

   How to scale without attributing everything | Vuk Vukosavljevic, HeyReach | Measurement Layer EP02

   Most B2B SaaS marketers try to attribute every conversion. Vuk Vukosavljevic, CMO at HeyReach, thinks that's a losing game. HeyReach finished 2025 at $10M ARR and is now at $16M, with 90%+ of revenue coming from inbound. Before HeyReach, Vuk led marketing at Lemlist and helped scale that company to $13M+ ARR. Both companies sell outbound tools. Both grew through inbound. And Vuk built the systems that made that scaling work. Without trying to attribute every touchpoint along the way. In this episode, Vuk explains: - Why he refuses to attribute every conversion, and what he tracks instead - How HeyReach built a word-of-mouth engine that drives 90% of new revenue - The three-layer distribution model: user ambassadors, creators, and personal brands - How to measure qualified free trials vs total free trials (5,000+ signups per month) - When to trust dashboards vs when to trust customer conversations - How he uses AI to analyze sales calls, support tickets, and LinkedIn engagement at scale - Why marketing and sales attribution don't matter as long as the scoreboard moves - The one measurement change every SaaS marketing leader should make next week Timestamps: (00:00) Intro (00:33) What Vuk owns at HeyReach and how growth is structured (01:38) The state of measurement Vuk inherited vs built (04:08) HeyReach's word-of-mouth machine: the three-layer model (07:41) When someone converts as "organic search": what actually happened (12:32) SaaS revenue attribution: sales-led vs self-serve, T1/T2/T3 tiering (16:43) Balancing CAC decisions today vs LTV data six months later (20:35) How to measure signals that don't show up in dashboards (24:05) Metrics his team deprioritized: qualified vs total (26:05) Vuk's full measurement framework in one summary (29:19) Dashboards vs conversations: why you need both (33:39) The one change every SaaS marketing leader should make next week Find Vuk: https://robustmedia.agency/ (https://robustmedia.agency/) Measurement Layer is a podcast where I sit down with practitioners and unpack how marketing measurement actually gets built. New episode every month on YouTube, Spotify, and Apple Podcasts. Host: Konstantin Vashkevich. Ten years working on one problem: making marketing measurement actually work. Currently Head of Marketing at RedTrack.io (http://RedTrack.io). Find me: https://www.linkedin.com/in/konstantin-vashkevich/ (https://www.linkedin.com/in/konstantin-vashkevich/)

6. 

   30 Jun 2026

   EP01 — Adrian Ćwikła: Tracking B2B marketing in a global BPO

   Most B2B marketing teams think they need more tools. Adrian Ćwikła thinks the opposite. Adrian is Head of Digital Marketing at Conectys — a global B2B customer experience outsourcing company operating across EMEA, APAC, and the Americas. In this conversation, he walks through how he tracks B2B marketing without enterprise complexity. The stack is simple: Google Analytics, Google Tag Manager, HubSpot. The discipline behind it is what makes it work. We cover: (00:00) Intro and what Conectys does (01:30) How a small marketing team operates across continents (04:15) The full tracking stack: GA, GTM, HubSpot (06:50) Why measuring B2B buyers is harder than measuring SaaS or e-commerce (08:40) The first thing Adrian changed when he stepped into the role (11:20) A landing page experiment that looked like a failure but converted 15 months later (14:20) Where AI saves time and where Adrian pulled back (17:50) How marketing data shifts perception of marketing as a cost center (21:00) "Data is a language" — talking to decision makers (25:50) The one change a B2B head of marketing should make next week About the host: Kostya Vashkevich. More than 10 years working on one problem: making marketing measurement actually work. Currently Head of Marketing at RedTrack.io (http://RedTrack.io). About the show: Measurement Layer is a podcast where I sit down with practitioners and unpack one specific piece of measurement infrastructure per episode. Operator-to-operator. No fluff. New episode every Tuesday. Find Adrian: https://www.linkedin.com/in/adrian-cwikla/ (https://www.linkedin.com/in/adrian-cwikla/) Find me: https://www.linkedin.com/in/konstantin-vashkevich/ (https://www.linkedin.com/in/konstantin-vashkevich/)

Contact Measurement Layer

Guest appearances

Books guests

Based on episode analysis; this does not confirm that the show is currently accepting guests.