AdTech TAM / Solutions Practice Hub | adtechsanto.fun 

TAM / Solutions Engineering

Practice the client conversations, proofs, and fixes you ship every day.

Follow the steps, grab a proof, and ship a customer-ready note without hunting through docs.

01

Frame it

Identify the signal: bidstream, tag, video, or consent. Write the one-line problem.

02

Prove it

Collect one proof: HAR, bid log, tracker status, or dataLayer screenshot.

03

Send it

Write a 3-line update: what changed, what you proved, what you need next.

Your weekly path

Use this as the rhythm: read, try, prove, and store artifacts.

Day 1

OpenRTB and bidstream

Read imp, device, user, regs. Decode 5 logs and mark missing fields.

Proof: one annotated log with highlights.

Day 2

Tags and consent

Check dataLayer order, consent flags, and tag firing rules.

Proof: HAR before and after consent toggle.

Day 3

Video and CTV

Validate VAST/VMAP, companions, and SSAI responses.

Proof: screenshot of Impression 200 + one SSAI status.

Day 4

Header bidding

Run Prebid client and server; compare timeout and win rate.

Proof: CSV of bidder timings and wins.

Day 5

Customer updates

Draft RCAs and partner emails with the proofs you collected.

Proof: one 3-paragraph RCA with attachments.

Focus areas for TAM / Solutions

Each row = what to know, what to do, what to show the customer.

Focus Area Topics Practice Tasks
Bidstream readiness OpenRTB 2.6, imp/device/user/regs, bidfloor, no-bid reasons Decode 20 logs; mark missing fields; draft a red-flags checklist
GAM delivery fixes Key-values, priority, unified auction vs waterfall, pacing Mock a network; fix 5 delivery issues; capture before/after screenshots
Video/CTV QA VAST/VMAP trackers, companions, SSAI errors, ad pods Validate 5 VAST tags; hit Impression tracker; note one SSAI error and fix
Header bidding Prebid client vs PBS, timeouts, price granularity, adapter config Add 2 bidders; compare timeout impact; export console timings
Deals & SPO PMP/PG setup, seat mapping, deal troubleshooting, SPO signals Write a 4-line partner email template; triage 5 fake deal failures
Consent & privacy TCF/US privacy flags, consent mode, dataLayer order Toggle consent in Tag Lab; prove tags stay silent until consent is true
Data & APIs Reporting APIs, webhooks, KPIs (fill, win rate, viewability) Pull one API report; calculate fill and win rate; share a 3-line readout
Customer updates RCA framing, risk/impact, next steps, proof attachments Create a 3-paragraph RCA template; attach one log and one screenshot

Customer scenarios to role-play

Use these to practice calm, evidence-first responses.

Deal under-delivering

Checklist: seat ID correct, key-values present, floor vs bid mismatch, geo/device filters, recent no-bid reasons.

Action: pull 5 no-bid logs and send a 4-line update with findings and asks.

Video tracker missing

Checklist: Impression/start/firstQuartile hit 200, companions returned, player firing events, SSAI stripping trackers.

Action: click the tracker in Video QA; paste status and response time in the ticket.

Timeout complaints

Checklist: hb_timeout vs bidder latency, creative load time, CDN region, PBS vs client split.

Action: run auctions at 400 ms and 800 ms; chart win rate change.

Consent dispute

Checklist: consent flag set before GTM load, analytics_storage, ads_storage, geo override.

Action: record one HAR before consent and one after; highlight blocked vs allowed calls.

Artifacts to keep

Build a mini portfolio that doubles as customer-ready proof.

Bidstream starter kit

glossary.md, red-flags checklist, 5 annotated bid logs.

Video/CTV QA pack

vast-samples/, tracker-status.csv, ssai-errors.md, vmap.xml.

HB timing report

hb-test-page.html, timeout-report.md, pbs-vs-client.csv.

Customer update templates

3-paragraph RCA, partner ask template, success proof screenshots.

Explain it like a customer call

Plain language, one example, one proof to collect.

OpenRTB (postcard)

What: A postcard with ad slot, device, user, and rules.

Example: "Slot A, 300x250, Chrome on Android, city = Mumbai."

Proof: Circle imp, device.ua, user.id in one log.

VAST (recipe)

What: A recipe telling the player which file and which trackers to call.

Example: "Play MP4, then call start and firstQuartile URLs."

Proof: Paste a VAST tag in the validator and screenshot the 200 on Impression.

GAM (timetable)

What: A schedule of which line item runs first.

Example: "Priority 4 for sports pages; fallback house line."

Proof: Two-line mock: targeted plus fallback with the key-value you used.

Prebid (vote)

What: Everyone bids; highest wins before timeout.

Example: Two bidders, 800 ms; bidder A wins at 1.20.

Proof: Run local Prebid with 2 adapters and screenshot the timeout log.

SSAI (read aloud)

What: Server stitches ads; player streams one file.

Example: HLS manifest with EXT-X-CUE markers.

Proof: Open an HLS URL and find "CUE" in the manifest text.

Logs (crumb trail)

What: Each request leaves crumbs that explain what happened.

Example: "204 no content" or "403 blocked by policy."

Proof: Group 10 logs by status code and write one line per group.

Privacy and consent

What: Green = send marketing tags; red = block; yellow = limited.

Example: analytics_storage: denied until the user clicks Accept.

Proof: Toggle consent in Tag Lab and note how the payload changes.

Quick questions and answers

Customer-facing, evidence-first responses. Copy them into your notes.

What do I check first in a bid request?

Look at imp[].bidfloor, device.ua, regs.gdpr, and one user.id. If any are missing, note it and ask for a new log.

Try it: open one saved bid log and highlight these four fields.

Why is a VAST ad not firing?

Check that the Impression URL is reachable and that the player fires start and firstQuartile. If trackers return 404, the ad will look broken.

Try it: paste the VAST URL in the validator and click the first tracker.

How do I choose client vs server Prebid?

Client is faster to prototype; server is better when page weight matters or you need more bidders. Start client, then move to server when CLS or TTFB rises.

Try it: run one page with two client bidders; note page weight and timeout.

What do I ask a partner when deals fail?

Ask for: deal ID, expected seat, targeting rules, and the last 5 no-bid reasons. Share your ad unit codes and time range.

Try it: draft a 4-line email template for deal debugging.

Reference docs to open fast

Bookmark these. They are the "authority" links used in interviews and real fixes.

Questions? Drop them in your notes and answer with one log, one screenshot, or one code block.

Topic detail guides

Each card tells you what to read and how to practice. Save one explainer and one checklist per card.

OpenRTB Deep Dive

Read: OpenRTB 2.5/2.6 spec (imp, device, user, regs, source). Grab one SSP bid sample.

Do: Decode 50 bid logs; list 10 malformed fields; write a field meanings glossary.

Keep: glossary.md, issues-log.csv, decoded-bid.json.

Open detail page ->

Ad Serving + GAM

Read: GAM unified auction vs waterfall, key-values, line item priority, dynamic allocation.

Do: Mock a GAM network; set targeting; solve 10 pacing and delivery issues; document fixes.

Keep: gam-troubleshooting-sheet.md, export of line items, RCA notes.

Open detail page ->

VAST and CTV Basics

Read: VAST 4.1, VMAP, IAB SSAI notes; required tracking events and ad pods.

Do: Parse 10 VAST tags; build one valid VAST XML; find missing events; debug 5 CTV manifests.

Keep: vast-checklist.md, vmap.xml, event-audit.csv.

Open detail page ->

Prebid.js / PBS

Read: Prebid auction flow, price granularity, timeout docs; PBS workflow and latency.

Do: Run Prebid locally; add two bidders; compare client vs server timing; chart timeout impact.

Keep: hb-test-page.html, timeout-report.md, pbs-vs-client.csv.

Open detail page ->

Advanced CTV/SSAI

Read: HLS/DASH with stitched ads, SSAI error codes, podding logic.

Do: Debug 10 SSAI responses; inspect manifests; create VMAP with pods and sequencing.

Keep: ssai-debug-log.md, vmap-pod.xml, manifest-notes.txt.

Open detail page ->

Log Analysis at Scale

Read: Parquet and JSON basics; Athena/BigQuery query patterns for latency and errors.

Do: Analyze 500 logs; group latency spikes; classify top 10 error signatures.

Keep: queries.sql, latency-histogram.png, log-playbook.md.

Open detail page ->

Reference Library

Read: One-stop list of specs, platform docs, and data tooling references.

Do: Bookmark and pair each link with a short explainer in your notes.

Keep: references.md with links and takeaways.

Open references ->

4-week learning plan

Follow this sprint-style plan; ship artifacts weekly.

Week 1: Bidstream

  • OpenRTB decode 50 logs
  • Write field glossary
  • Red-flags checklist

Week 2: Ad Serving

  • Mock GAM network
  • Fix 10 delivery issues
  • Publish troubleshooting sheet

Week 3: Video/CTV

  • Build VAST + VMAP
  • Debug 5 CTV manifests
  • SSAI error cheat sheet

Week 4: HB and Logs

  • Prebid client vs PBS timing
  • Analyze 500 logs (latency/errors)
  • Ship log-playbook + timeout report