An app team has purchase events arriving in TikTok and an agency proposal to switch to value optimization. The setting is available. The unresolved question is whether the values represent the business well enough to authorize that switch. One qualifying purchase can unlock a feature without establishing a reliable basis for a revenue target.
TikTok app value-based optimization (VBO) uses purchase or in-app ad-revenue value to seek higher-value users, rather than optimizing only for completion of an event. Adopt it when the exact campaign path is eligible, the value feed is trustworthy and the reporting horizon matches the decision. Choose Highest Value or Target ROAS around the business constraint; neither strategy guarantees profit, complete budget delivery or better lifetime value.
This guide is for app founders, heads of growth and UA leads evaluating an agency recommendation or a bounded VBO test. Official TikTok documentation was checked on November 14, 2026. The acceptance framework and fictional example below are editorial analysis, not Sharply Labs client results.
What changes when an app moves from events to value?
An event-count goal asks whether a defined action happened. A value goal also needs an amount whose meaning is consistent across the events being used. Two purchases can count equally as actions while carrying different values. That makes value optimization a measurement commitment as well as a campaign choice.
TikTok’s current app VBO overview describes purchase value and in-app ad-revenue value under App Promotion, with Highest Value and Target ROAS bidding. Its statement of product intent is not evidence that a particular app will acquire more profitable users.
Before discussing bids, ask which value the business wants to influence. A subscription purchase amount, publisher ad revenue and net contribution are different quantities. If the campaign receives one while management evaluates another, record the relationship and its limits. Do not rename the received amount “LTV” simply because a platform description discusses long-term value.
Keep the earlier channel decision separate. The TikTok App Promotion readiness guide addresses whether the app has the proposition, creative and event readiness to test TikTok at all. This article addresses the narrower decision to commission value-based delivery once that foundation exists.
Separate interface eligibility from commercial readiness
TikTok’s October 2026 setup instructions list different prerequisites for purchase and ad-revenue value. For purchase value, the page specifies at least one unique purchase conversion event carrying value within the previous 365 days on TikTok or TikTok Ad Network. For ad revenue, it requires real-time postbacks through an MMP and adds SDK and detected value/currency requirements for the described Day 7 path.
Those are documented access conditions. They are not a recommendation to fund a campaign based on one observation. A usable test also needs enough relevant evidence to define the value, understand its variation and decide what a result would mean. There is no universal sample size supplied by this editorial framework.
Ask the operator to record the actual app, operating system, region, objective, optimization goal, value type, bid strategy and available horizon. An agency should show the campaign creation path it intends to operate. A screenshot from a Web campaign or another app does not establish eligibility for this one.
When documentation and the interface differ, preserve both records and resolve the discrepancy before approving a requirement-dependent launch. Do not interpret an unavailable option as a reason to substitute another value event silently. A purchase-led test and an ad-funded test may answer different questions even if both display a ROAS column.
Write a value contract before selecting a target
The following register is an editorial acceptance tool. It connects the signal entering delivery to the business decision waiting on it.
Boundary What the team records Reason to defer the test --------- Event meaning Action, trigger, production environment and responsible owner Test events or ambiguous triggers remain mixed in Amount Value source, currency, units and transformation The same number can mean different amounts Delivery Sender, destination, mapping and receipt evidence Sending is mistaken for accepted usable value Coverage Included products, markets and known omissions A partial feed is presented as the whole business Time Cohort start, value horizon and extraction date Immature and mature results are compared directly Economics Deductions and costs used for the business decision Attributed revenue is treated as contribution or cash
For purchase-led apps, review representative production events against the transaction evidence available to the team. Include low and high amounts, different products and any relevant currency paths. The purpose is to find semantic or transformation errors, not to upload private customer records into a marketing discussion. Redact examples and agree access separately.
For ad-funded apps, trace the publisher-revenue source through the measurement partner to TikTok. TikTok’s MMP ad-revenue instructions specify supported partner scope and real-time impression-level revenue postbacks. Verify the current integration for the actual partner; a general analytics connection is not sufficient proof of this path.
The ad revenue attribution guide covers collection, assignment and reconciliation in more detail. Keep those checks attached to the bidding proposal. A revenue report that looks plausible can still omit a material monetization source or contain values with a different revision status.
Highest Value versus Target ROAS: which constraint matters?
TikTok’s app VBO bidding comparison distinguishes spend-based Highest Value from goal-based Target ROAS. The business decision is how much flexibility it can allow while learning about value, not which label sounds more advanced.
Decision Highest Value may fit when Target ROAS may fit when --------- Immediate question What value can this bounded spend opportunity produce? Can delivery operate around a defined return goal? Business flexibility The team accepts uncertainty in observed return A meaningful return constraint is already defined Evidence prerequisite Value is reliable enough to investigate Value and the target’s denominator/horizon are understood Review risk Spending is mistaken for commercial success A target entered in the interface is mistaken for a guarantee
This table interprets the documented distinction for procurement. It does not establish that either strategy is available for every app or that one will outperform the other. Verify compatibility in the account before commissioning execution.
A target should have a written definition. Specify which purchase or ad-revenue measure it uses, which horizon it covers and which business costs remain outside it. A finance hurdle expressed in net contribution cannot simply be pasted into a platform field using gross attributed value. First establish the bridge between those measures.
The dedicated Target ROAS page describes its Smart+ App scope and placement limitations. Do not use a generic product summary to promise identical inventory or controls across campaign paths. Campaign automation is a separate operating decision, covered in the Smart+ versus manual app comparison.
Day 0 and Day 7 describe value maturity, not just patience
TikTok’s overview distinguishes install-day ad revenue from ad revenue accumulated within the first seven days after install. A longer horizon changes what can enter the numerator. It also means that a recent cohort has not yet supplied the complete observation the decision requires. Calendar days since a campaign launch and maturity of each acquired cohort are different clocks.
Do not assume every purchase/ad-revenue, OS and bidding combination supports the same horizon. The current bidding matrix lists combinations separately, while broader setup language describes Day 0 and Day 7 choices. Read the specific matrix alongside the actual account rather than inferring universal Day 7 purchase support.
Preserve the metric definition when exporting. TikTok’s in-app metric instructions explain how to select app columns, including Day 0 purchase and ad-revenue ROAS. An unlabeled “ROAS” screenshot is a weak handoff because the next reviewer cannot tell what was measured.
For a comparison, use equally mature cohorts, consistent value definitions and a documented extraction time. If part of the observation is unavailable, leave it unavailable. Do not fill a missing Day 7 result by multiplying Day 0 value by an unexplained factor. Forecasting can be useful, but assumptions and observed results need separate columns.
A fictional example: the longer window changes the approval question
Consider two fictional campaign packages, each spending $5,000. Package A produces $2,000 of recorded install-day value; package B produces $2,500. On that definition, observed Day 0 ROAS is 40% for A and 50% for B. These are invented figures to illustrate the calculation, not benchmarks or expected TikTok results.
Now suppose equally mature records show $4,500 of seven-day value for A and $4,000 for B. Day 7 ROAS becomes 90% and 80%, respectively. The ordering reverses because the observation horizon changes. That does not establish causality, statistical reliability or the best bidding strategy. The packages may differ in other ways.
Next suppose the finance review uses contribution after the relevant revenue deductions and variable costs. Its numerator may be lower still. Neither the Day 0 nor Day 7 ratio alone answers whether the team can fund another cohort. The mobile app ROAS guide explains why revenue return, contribution and cash payback require different definitions.
The agency’s recommendation should therefore name the commitment under review. Continuing a small diagnostic test, buying a larger creative package and expanding market spend are different decisions. Evidence sufficient for one may be insufficient for another. A useful report says which action the observations support and what remains unknown.
If reliable package-level downstream records are unavailable, do not allocate total company revenue between packages in proportion to installs and call the result measured ROAS. That would introduce an assumption precisely where the decision needs evidence. Record the gap, its owner and whether a narrower test can still serve a useful purpose.
Use three acceptance tests before authorizing VBO spend
First, accept the signal. The event and amount have a documented meaning, transformations are understood, representative records reconcile and known omissions are disclosed. Acceptance means the team understands what it is asking delivery to pursue. It does not mean every future event is guaranteed correct.
Second, accept the campaign boundary. The proposed path is available for the app, value type and horizon. Budget authority, creative permissions and account responsibilities are explicit. If access depends on an integration change, that work has an owner and a verification step before media is committed.
Third, accept the decision rule. The reviewer knows which mature evidence could justify continuing, repairing, stopping or commissioning another test. Define the review horizon, exposure limit and escalation conditions before seeing a flattering return chart. Avoid an arbitrary universal “successful ROAS” threshold.
These tests help distinguish a delivery problem from a signal problem. If value receipt breaks, preserve the settings and relevant evidence, investigate collection and mapping, and assess whether further spend can still answer the approved question. If signals are healthy but delivery is limited, review the actual constraints rather than rewriting the revenue definition to improve a report.
If delivery is healthy but business value disappoints, the next question may concern product fit, creative promise or economics. Changing a bid target cannot itself establish the cause. Keep a dated change log so the team can tell which intervention preceded which observation without claiming that sequence proves causality.
What should the agency proposal deliver?
Ask for an eligibility record, the value contract, a measurement readout specification and a scoped operating plan. Each deliverable should name the owner, access dependency and acceptance evidence. “Launch VBO” is too vague to tell procurement whether the fee covers event investigation, engineering coordination, creative work or only campaign operation.
The operating plan should distinguish media budget from service fees and production costs. It should state what happens when the value feed becomes unreliable, an expected control is absent or the observation period remains incomplete. Those are ordinary operating conditions to plan for, rather than reasons to invent a guaranteed result.
Request a recommendation that includes disqualifying conditions. A team with unresolved currency semantics may need measurement repair first. A team with a reliable feed but no agreed commercial hurdle may need a bounded exploratory decision. A team with neither access nor implementation ownership should avoid committing to a launch date dependent on both.
Sharply Labs’ mobile app growth services explicitly include reviewing TikTok event and value objectives against account requirements, creative readiness and event quality. The broader growth service scope provides the commercial context. Specific account access, engineering changes and ongoing operation require an agreed engagement; this guide does not imply privileged platform access or private client evidence.
Bring the value definition and the decision waiting on it
For an established app team considering TikTok VBO, bring a redacted campaign-path record, representative value examples, the reporting horizon and the budget decision currently blocked. Include known omissions and the business definition of an acceptable outcome.
An initial Sharply Labs conversation can identify the evidence gaps, responsible owners and a prioritized readiness agenda. Detailed account auditing, integration work, experiment design and campaign operation are separately scoped. The conversation does not promise eligibility, a completed audit, higher ROAS, profitable acquisition or a particular learning timeline.