Approach

Audit first. Then fix the measurement. Then spend.

In that order, every time, because the alternative is optimising toward a number nobody has checked. Goloporim engagements start with a fixed-scope audit that produces a document you keep whether or not you continue — which is also the cheapest way for you to find out whether we are any good.

Engagement model

What happens, and when.

  1. 01

    Week 0

    Written brief, then a call

    You send the apps, the stores, the MMP and roughly what you spend. We reply in writing with what we would look at first and what it would cost to look at it. The call happens after that document exists, not instead of it.

  2. 02

    Weeks 1–4

    Fixed-scope audit

    Access and inventory, then an end-to-end trace of a real install on both platforms, then reconciliation against the store consoles, the ad platforms and your warehouse. Ends in a written findings document with the fixes ordered by what they cost you.

  3. 03

    Week 5

    Scope and contract

    If you want us to execute, we scope it: workstreams, named deliverables, cadence, term, notice period and price, in USD, in writing. If you only wanted the audit, the engagement ends here and the document is still yours.

  4. 04

    Weeks 5–8

    Fix the measurement

    Event taxonomy, server-to-server events, conversion-value schema, source identifier structure, deep links, fraud rules, raw-data exports. We do not touch bids while the baseline is still moving.

  5. 05

    Weeks 9–12

    Operate and test

    Channel and creative tests sized to reach a conclusion, store-side experiments inside each platform’s concurrency limits, and a weekly decision log against the now-fixed baseline.

  6. 06

    Quarterly

    Re-audit and re-plan

    Platforms change — a new SKAN behaviour, a new placement, a new store surface. Every quarter the measurement gets re-checked and the plan gets rewritten rather than extended by inertia.

What we need from you

The access list, published, so procurement can start before the first call.

Nothing here is unusual and nothing here is more than the work requires. We will sign your NDA, your DPA and your security questionnaire; send them with the brief and we will return them before the kickoff.

MMP
Admin on the account, or a user with configuration rights plus raw-data export access. Admin is needed to change schemas and fraud rules; read-only is enough for the audit itself.
Ad accounts
Standard or admin access on Apple Ads, Google Ads, Meta, TikTok and any SDK networks in the plan — on your own accounts, in your own billing.
App Store Connect
Marketing or Analytics role for App Analytics, product pages and Product Page Optimization.
Google Play Console
Store listing and experiment permissions, plus access to acquisition reports.
Product analytics
Read access to whatever you already use — Amplitude, Mixpanel, Firebase, an in-house pipeline. We do not need to replace it.
Warehouse
Read access to the tables that hold revenue, or an export we can reconcile against. This is the step that makes the finance bridge possible.
Engineering contact
One named person who can ship an SDK or event change, and who is in the room when the schema is decided.

Reporting

Cadence, and who it is for.

Reporting is a decision record, not a performance. If a week produced no decision, the update says so in one line.

Weekly

A written update: what changed, what it did, what is next, what is blocked. Fifteen minutes to read, no slides.

Weekly call

Thirty minutes, optional, with your growth owner. Skipped when there is nothing a document cannot carry.

Monthly

Portfolio review: channel performance against plan, creative conclusions, store-side test readouts, and the reconciliation to finance.

Quarterly

Re-audit, re-plan, and an honest account of what we got wrong in the previous quarter.

Boundaries

What we will not do.

Published, because every one of these has been asked for at least once, and because an agency that will not tell you its limits in advance is telling you something anyway.

We will not guarantee a result

Media outcomes depend on your product, your price and your market, and most of that is not ours to move. Anyone who guarantees a lift has either priced the guarantee into the fee or is not planning to honour it.

We will not take a rebate or a markup on your media

You pay the platforms directly, on your own cards, in your own accounts. We are paid by you, which is the only way our advice about where to spend can be worth anything.

We will not buy incentivised or rebrokered traffic

No incent, no motivated installs, no sub-brokered inventory dressed as a direct network. It flatters an install number and poisons every cohort downstream.

We will not hold your accounts

Ad accounts, MMP contract, store accounts and raw data stay in your name from day one. Leaving us should cost you a vendor, not your history.

We will not report a number we cannot reconcile

If the MMP and the store console disagree, you get both and the reason, not whichever one looks better.

We will not take an account we cannot help

If the problem is retention, pricing or product, we will say so in the audit and tell you not to spend more until it is fixed.

Next step

Bring us the account you are not sure you can trust the numbers on.

Tell us the apps, the stores, the MMP and roughly what you spend. We will reply with what we would look at first, in writing, before anyone asks you to sit on a call.