Pricing model operations

CPI Advertising: Campaign Setup, Bidding and Optimization

Plan CPI advertising with clear billing units, tracking, source controls, creative tests, landing-page checks and scale rules.

Billable unitattributed install
Base formulamedia spend divided by attributed installs
Decision metricCPI beside activation rate, retention, payer conversion and lifetime value
Best fitmobile app acquisition when attribution, deep links, install validation and post-install events are implemented
CPI Advertising: Campaign Setup, Bidding and Optimization operating model
Answer first

What cpi advertising should mean in a real campaign

CPI Advertising requires a declared billable event, a campaign objective and a written optimization rule. Setup decisions should make it possible to trace spend from the auction or delivery event through the landing page and into the mature business outcome without losing source, creative or device context.

The first cpi advertising document should state the billable event, the formula, the attribution window and the accepted business outcome. For this model, the billable unit is attributed install, and the base formula is media spend divided by attributed installs. The formula is only the starting point. The commercial decision should use cpi beside activation rate, retention, payer conversion and lifetime value after the underlying outcomes have had enough time to mature.

Use media source, campaign, app version, operating system, geo, device and install cohort as the minimum reporting breakdown. The central risk is buying inexpensive installs that never activate, retain or create enough value to recover acquisition cost. A source-level structure, a maximum test loss and a reason-coded change log prevent the team from interpreting a temporary average as a durable result. In this cpi advertising workflow, the practical reason for this control is to operate cpi advertising as a measured campaign with explicit billing, attribution and optimization rules while preserving a source-level explanation for every material change.

Operating controls

Six layers that make cpi advertising measurable

The pricing label becomes useful when billing, source quality, tracking and scale rules are explicit.

01

Billing definition

Document exactly when a attributed install is counted, filtered, adjusted and billed. The page should distinguish the configured bid, the effective price and the cost that remains after invalid-event or reconciliation adjustments. For cpi advertising, connect this layer with the declared objective and keep the decision reversible until the result matures.

02

Break-even value

Calculate the maximum affordable media cost from accepted outcome value, variable costs, rejection or reversal rates and required margin. Use cpi beside activation rate, retention, payer conversion and lifetime value as the commercial decision layer. For cpi advertising, connect this layer with the declared objective and keep the decision reversible until the result matures.

03

Source transparency

Preserve media source, campaign, app version, operating system, geo, device and install cohort. Source and placement detail lets the team stop waste without discarding the entire model or hiding weak inventory inside a blended account average. For cpi advertising, connect this layer with the declared objective and keep the decision reversible until the result matures.

04

Tracking chain

Carry campaign, source, creative and event identifiers through the landing path. Reconcile platform delivery with clicks, attributed installs, first opens, activation, retention, revenue, uninstall and fraud signals before changing bids or declaring a winner. For cpi advertising, connect this layer with the declared objective and keep the decision reversible until the result matures.

05

Creative and page fit

On this CPI Advertising: Campaign Setup, Bidding and Optimization page, Creative and page fit matters because it changes what the advertiser should verify before committing budget or operating effort. Review Match, promise, destination, paid, unit and attractive together, because a strong result in one of them should not conceal a material failure in another. Use the finding to choose a specific action—keep, cap, exclude, renegotiate, retest or stop—rather than recording a score with no operational consequence. For a FroggyAds campaign, translate this conclusion into the narrowest applicable targeting or budget change and reconcile the result with the accepted business event.

06

Scale governance

Use written stop, revise and scale rules. Increase budget only after the result repeats, the outcome window matures and the next increase remains below the declared break-even limit. For cpi advertising, connect this layer with the declared objective and keep the decision reversible until the result matures.

Connect the guide to live testing

Connect CPI Advertising to a controlled audience test

Use the choices established in “Six layers that make cpi advertising measurable” to define one audience, budget and source set in FroggyAds. Keep the surrounding offer and measurement rule stable so the test adds evidence to cpi advertising instead of mixing several changes at once.

Create My Free Account
Illustration of audience targeting controls for a cpi advertising test
Formula and forecast

Translate the paid unit into a break-even range

On this CPI Advertising: Campaign Setup, Bidding and Optimization page, Translate the paid unit into a break-even range matters because it changes what the advertiser should verify before committing budget or operating effort. Translate the section into checks for Start, forecast, accepted, Subtract, fulfillment and sales; this keeps the recommendation tied to the page's real task instead of generic marketing language. Keep the baseline unchanged while testing the next hypothesis; that comparison is what makes the decision reproducible. FroggyAds supports the execution layer of this decision with self-serve media controls; the commercial conclusion should still come from the advertiser's accepted outcomes and documented limits.

The practical role of Translate the paid unit into a break-even range in CPI Advertising: Campaign Setup, Bidding and Optimization is to expose the exact condition that can change the buyer's next action. Keep the review anchored to rate, forecast, range, promise, Competition and device; those details are the parts of this section that can materially change the recommendation. When the evidence is strong, carry the exact setting or requirement into the next campaign step instead of broadening several variables at once.

Separate configured values from effective values. A bid ceiling, target or advertised minimum is not necessarily the amount paid. Automated products can adjust delivery or bids within platform-specific rules. The useful report shows the actual cost, the paid denominator and clicks, attributed installs, first opens, activation, retention, revenue, uninstall and fraud signals for the same cohort. In this cpi advertising workflow, the practical reason for this control is to operate cpi advertising as a measured campaign with explicit billing, attribution and optimization rules while preserving a source-level explanation for every material change.

Paid unitattributed install
Base formulamedia spend divided by attributed installs
Primary business metricCPI beside activation rate, retention, payer conversion and lifetime value
Required reporting splitmedia source, campaign, app version, operating system, GEO, device and install cohort
Maturity evidenceclicks, attributed installs, first opens, activation, retention, revenue, uninstall and fraud signals
CPI Advertising: Campaign Setup, Bidding and Optimization pricing decision matrix
Implementation workflow

A seven-step cpi advertising operating process

Use a bounded sequence so the first budget creates evidence rather than a collection of unrelated changes.

01

Define the billable event

Write the exact attributed install definition for cpi advertising. Include validation, view or click thresholds, attribution, time zone and any platform-specific adjustments. The cpi advertising work log should state the evidence required before the next step begins.

02

Model the economics

Use media spend divided by attributed installs for the paid unit, then translate that result into cpi beside activation rate, retention, payer conversion and lifetime value. Include non-media costs and a margin reserve. The cpi advertising work log should state the evidence required before the next step begins.

03

Instrument the path

Test redirects, landing pages, conversion events, postbacks and source parameters. A pricing-model test is not ready while the paid event and business outcome cannot be reconciled. The cpi advertising work log should state the evidence required before the next step begins.

04

Launch a bounded cell

Choose one offer, a limited GEO and device scope, a small creative set and a maximum test loss. Preserve media source, campaign, app version, operating system, geo, device and install cohort from the first paid event. The cpi advertising work log should state the evidence required before the next step begins.

05

Wait for maturity

Separate provisional and mature results. For cpi advertising, do not compare cohorts that have had different time to convert, be approved, generate revenue or reverse. The cpi advertising work log should state the evidence required before the next step begins.

06

Apply reason-coded actions

Mark each change as bid, creative, source, targeting, page, tracking or policy. Record the previous value and the expected effect so the next review can test the hypothesis. The cpi advertising work log should state the evidence required before the next step begins.

07

Scale with a control

Keep a stable control while increasing spend on proven cells. Watch whether effective cost, source mix, frequency, outcome quality or margin changes as the campaign reaches more inventory. The cpi advertising work log should state the evidence required before the next step begins.

CPI Advertising: Campaign Setup, Bidding and Optimization implementation workflow

Choose the execution format

Choose a paid-media format that supports CPI Advertising

Use the criteria around “A seven-step cpi advertising operating process” to decide whether push, native, display or pop fits the message and destination. Set format, targeting and spend as campaign controls in FroggyAds while the cpi advertising decision remains the standard for judging the result.

Create My Free Account
Illustration comparing advertising formats for cpi advertising execution
Measurement design

Reconcile delivery, analytics and accepted value

The headline metric for cpi advertising is cpi beside activation rate, retention, payer conversion and lifetime value. Define its numerator, denominator, currency, time zone, attribution rule and maturity window before comparing campaigns. Platform delivery, analytics sessions, conversion events, CRM status and collected revenue can settle at different times.

For CPI Advertising: Campaign Setup, Bidding and Optimization, the Reconcile delivery, analytics and accepted value checkpoint should answer a concrete buyer question rather than repeat a generic framework. Compare Build, reconciliation, table, connects, spend and paid under the same scope and review window; if one is unknown, keep that uncertainty explicit rather than filling the gap with an estimate. Keep the baseline unchanged while testing the next hypothesis; that comparison is what makes the decision reproducible. A controlled FroggyAds test can turn this section into measurable evidence: keep the conversion definition stable, preserve source identifiers and compare marginal performance before expanding.

Read early diagnostics without promoting them to final outcomes. Click-through rate, completion rate, viewability, page engagement and raw conversion rate can explain where the path breaks. The budget decision should wait for clicks, attributed installs, first opens, activation, retention, revenue, uninstall and fraud signals to mature. In this cpi advertising workflow, the practical reason for this control is to operate cpi advertising as a measured campaign with explicit billing, attribution and optimization rules while preserving a source-level explanation for every material change.

Traffic and network evaluation

Choose inventory by transparency and control

Volume claims and headline rates cannot replace source-level evidence.

A platform used for cpi advertising should expose the billable event, reporting latency, source or placement identifiers, targeting controls, invalid-event treatment and conversion-tracking options. Check whether the account can separate discovery traffic from proven sources and whether changes are available at the level where performance actually differs.

Ask how cost per install is implemented for the chosen format. The same label can describe different auction, validation or optimization rules across platforms. For automated variants, document the maximum bid or target, the signals used, the learning period and the advertiser controls that remain available. In this cpi advertising workflow, the practical reason for this control is to operate cpi advertising as a measured campaign with explicit billing, attribution and optimization rules while preserving a source-level explanation for every material change.

Run the first cpi advertising test with a clear loss limit and a narrow question. Compare the platform report with your analytics and business records. A network deserves more budget when the differences are explainable, the quality controls work and the result survives a mature acceptance window.

Creative and landing experience

Make every paid event lead to the same promise

Within CPI Advertising: Campaign Setup, Bidding and Optimization, Make every paid event lead to the same promise should connect the page's stated intent to evidence that a media buyer or marketing team can actually inspect. Translate the section into checks for align, creative, landing, path, offer and eligibility; this keeps the recommendation tied to the page's real task instead of generic marketing language. Set a written pass condition and a rollback condition before acting, so the team can reverse the change without rewriting the history of the test. Where this leads to paid acquisition, FroggyAds gives you a self-serve campaign environment for applying the relevant targeting, budget and source controls while your own analytics verifies downstream value.

01

Promise

For cpi advertising, the ad should state one truthful benefit that the destination can verify.

02

Qualification

Use the message to attract the user who can complete the accepted outcome, not merely the cheapest attributed install.

03

Continuity

Repeat the core reason to act on the landing page so cpi advertising performance reflects the offer rather than surprise or confusion.

04

Speed

Test the destination on purchased devices and connections. Lost sessions distort effective CPI economics.

05

Proof

Use transparent terms, relevant evidence and realistic expectations. Fabricated urgency or reviews weaken both trust and measurement.

06

Tracking

Preserve source, placement, creative and event identifiers so the complete cpi advertising path remains attributable.

Put the guide into practice

Turn CPI Advertising into a bounded campaign test

With “Make every paid event lead to the same promise” documented, launch only the next reversible test. Set a spending limit, preserve the baseline and use source-level and audience controls so the next step depends on qualified outcomes for cpi advertising, not activity volume.

Create My Free Account
Illustration of a campaign launch checklist for cpi advertising
Decision scenarios

How to respond when cpi advertising metrics disagree

Use the disagreement to identify the broken layer instead of changing the entire campaign.

01

The paid rate falls but CPA rises

The cheaper attributed install may be coming from weaker sources, lower viewability, accidental response or a landing mismatch. Compare source-level qualified sessions and accepted outcomes before calling the lower rate an improvement. In a cpi advertising review, document the diagnosis and the single next change before editing the campaign.

02

Delivery grows while quality is flat

Expansion may have changed the inventory mix. Hold the best-performing cells stable, isolate the new sources and compare clicks, attributed installs, first opens, activation, retention, revenue, uninstall and fraud signals after the same maturity window. In a cpi advertising review, document the diagnosis and the single next change before editing the campaign.

03

One creative wins early

For CPI Advertising: Campaign Setup, Bidding and Optimization, the One creative wins early checkpoint should answer a concrete buyer question rather than repeat a generic framework. Translate the section into checks for Confirm, winner, preserves, accepted, benefiting and unequal; this keeps the recommendation tied to the page's real task instead of generic marketing language. When the evidence is strong, carry the exact setting or requirement into the next campaign step instead of broadening several variables at once.

04

Platform and analytics disagree

Check time zones, click IDs, view or click definitions, redirect loss, duplicate rules, consent and attribution windows. Do not average the systems together. Reconcile the event chain with reason codes. In a cpi advertising review, document the diagnosis and the single next change before editing the campaign.

05

The model works in one GEO only

Treat the GEO as a separate economic cell. Price, device mix, payment behavior, language and source availability can change the break-even point. Do not copy the bid into another market without a local test. In a cpi advertising review, document the diagnosis and the single next change before editing the campaign.

06

Scale reduces margin

A buyer evaluating CPI Advertising: Campaign Setup, Bidding and Optimization can use Scale reduces margin to make the page actionable: identify the condition, document the evidence, and define the response. Compare larger, budget, reaching, expensive, auctions and weaker under the same scope and review window; if one is unknown, keep that uncertainty explicit rather than filling the gap with an estimate. Use the finding to choose a specific action—keep, cap, exclude, renegotiate, retest or stop—rather than recording a score with no operational consequence.

Failure prevention

Eight mistakes that weaken cpi advertising

When using CPI Advertising, apply this rule only to the conditions and decision described on this page. Most pricing-model losses come from small definition, tracking and decision defects that survive because the blended account still looks acceptable. Use the checklist before launch and during every material budget review.

  1. 01Using a different CPI event definition in the platform, analytics and finance reports. Assign an owner, a reason code, a measurable correction and a review date. For Cpi Advertising, validate this point against CPI Advertising, Plan CPI advertising, cpi advertising work and keep it separate from the What Is Cpi Advertising intent.
  2. 02Comparing cpi advertising rates across GEOs, devices or formats without normalizing the denominator. Assign an owner, a reason code, a measurable correction and a review date.
  3. 03Changing bid, creative, source rules and landing page in the same optimization cycle. Assign an owner, a reason code, a measurable correction and a review date. For Cpi Advertising, validate this point against CPI Advertising, Plan CPI advertising, cpi advertising work and keep it separate from the What Is Cpi Advertising intent.
  4. 04Scaling provisional conversions before acceptance, retention or revenue has matured. Assign an owner, a reason code, a measurable correction and a review date. For Cpi Advertising, validate this point against CPI Advertising, Plan CPI advertising, cpi advertising work and keep it separate from the What Is Cpi Advertising intent.
  5. 05Judging cpi advertising from a blended account average that hides weak source cells. Assign an owner, a reason code, a measurable correction and a review date.
  6. 06Treating a lower rate as success while qualified sessions and accepted outcomes decline. Assign an owner, a reason code, a measurable correction and a review date. For Cpi Advertising, validate this point against CPI Advertising, Plan CPI advertising, cpi advertising work and keep it separate from the What Is Cpi Advertising intent.
  7. 07Allowing tracking loss, duplicate events or attribution differences to remain unexplained. Assign an owner, a reason code, a measurable correction and a review date. For Cpi Advertising, validate this point against CPI Advertising, Plan CPI advertising, cpi advertising work and keep it separate from the What Is Cpi Advertising intent.
  8. 08Keeping a losing cpi advertising segment active because the total campaign is still above break-even. Assign an owner, a reason code, a measurable correction and a review date.
30-day operating plan

Move from definition to a repeatable CPI decision

Treat Move from definition to a repeatable CPI decision as a specific gate for CPI Advertising: Campaign Setup, Bidding and Optimization, not as a reusable checklist item that means the same thing on every page. Use fixed, observation, window, spend, changes and follow as the traceable inputs for this section, then state which missing item would be serious enough to stop or narrow the decision. Do not scale the conclusion beyond the evidence window; repeat the check after the next meaningful change in volume, scope or audience. If the next step is a media test, FroggyAds lets the advertiser keep campaign settings and source-level performance visible instead of treating traffic volume as proof of success.

01

Days 1 to 3: define

Document the CPI event, formula, value model, attribution rule and maximum test loss for cpi advertising. Verify the destination and every measurement handoff before buying volume.

02

Days 4 to 10: launch narrow

Run a bounded cpi advertising cell with limited GEO, device, sources and creatives. Monitor delivery and obvious technical failures, but avoid rewriting the campaign before representative evidence arrives.

03

Days 11 to 20: reconcile

When using CPI Advertising, apply this rule only to the conditions and decision described on this page. Compare platform delivery with clicks, attributed installs, first opens, activation, retention, revenue, uninstall and fraud signals. Separate provisional and mature outcomes, remove repeated failures and keep a small controlled budget for source discovery.

04

Days 21 to 30: repeat or scale

Increase spend only where cpi beside activation rate, retention, payer conversion and lifetime value remains inside the target range. Keep the previous stable setup available and record how the larger auction footprint changes effective cost and source mix. In this cpi advertising workflow, the practical reason for this control is to operate cpi advertising as a measured campaign with explicit billing, attribution and optimization rules while preserving a source-level explanation for every material change.

Primary references

Standards and first-party evidence for CPI Advertising

The practical role of Standards and first-party evidence for CPI Advertising in CPI Advertising: Campaign Setup, Bidding and Optimization is to expose the exact condition that can change the buyer's next action. Compare standards, official, documentation, make, operating and your under the same scope and review window; if one is unknown, keep that uncertainty explicit rather than filling the gap with an estimate. If the evidence does not support the current assumption, narrow the scope or run the smallest reversible test that can resolve it.

Frequently asked questions

CPI Advertising FAQ

Answers focus on billing definitions, measurement, quality and responsible scaling.

When does CPI advertising fit an app acquisition plan?

It fits when an install is attributable, the app and store listing are ready, and the team can judge post-install quality. Paying per install is less useful when onboarding is broken or the business cannot distinguish valuable users.

For CPI Advertising, what should an initial CPI campaign learn before it seeks volume?

The first campaign should establish valid install cost, attribution reliability, creative response, and early user quality by source. A capped test gives the team evidence without making scale the condition for learning.

For CPI Advertising, how should CPI targeting reflect the app's actual use case?

Start with eligible markets, supported devices, language, operating system, and contexts linked to a real user need. Broad reach is not an advantage when people cannot use the product or understand its store page.

For CPI Advertising, what connects a CPI ad to a credible store-page experience?

The ad, screenshots, description, permissions, and first-run experience should present the same core value and conditions. Continuity reduces accidental installs and gives creative comparisons a fairer quality signal.

For CPI Advertising, which costs sit outside the reported CPI?

Creative production, platform and data fees, attribution, fraud review, store optimization, and campaign management may be separate. Include the costs relevant to the acquisition decision rather than relying on media CPI alone.

For CPI Advertising, how should an app team configure attribution for CPI ads?

Agree the install event, lookback window, reinstallation rule, source identifiers, privacy settings, postbacks, and rejection handling. Test the path on supported devices before paid delivery begins.

For CPI Advertising, what should a weekly CPI report show beyond installs?

Show spend, valid and rejected installs, source, creative, market, device, onboarding, meaningful early actions, and cohort timing. That view shows if lower cost brought genuinely usable acquisition rather than install volume alone.

For CPI Advertising, why can CPI improve while app growth weakens?

A campaign may attract cheaper but less engaged installers, expand into weaker placements, or count events before validation. Read CPI with post-install behaviour and cohort value before rewarding the apparent improvement.

For CPI Advertising, how can CPI campaigns limit fraudulent or accidental installs?

Use transparent sources, measurement-platform checks, frequency and placement controls, conservative caps, anomaly alerts, and post-install review. Pause questionable subsources promptly rather than averaging them into trusted delivery.

What proves that CPI advertising can scale responsibly?

Scaling is supported when attribution, valid-install cost, and useful post-install behaviour remain stable as a cap rises. Expand markets or sources separately and retain cohort reporting for the added volume.

Launch with evidence

Turn cpi advertising into a controlled campaign test

For CPI Advertising, treat this as a page-specific operating check rather than a universal benchmark. Start with one objective, a precise paid-event definition, transparent tracking, source-level controls and a written stop or scale rule. Results depend on the offer, creative, landing page, GEO, bid and optimization.

decision framework

Cpi Advertising: choose the billing model by measurable business value

Direct answer: Cpi Advertising should be evaluated by the exact billable event, inventory transparency, conversion tracking, source-level controls and the value produced after validation. A low headline rate is not automatically efficient. Compare qualified outcomes, not only the platform charge.

Define the event before bidding

On CPI Advertising, use this control to keep the page's evidence and action traceable. Write down what triggers a charge, which events count as qualified, how duplicates and invalid activity are handled, and which reporting window will be used. Keep the media metric separate from the commercial outcome. CPM measures impressions, CPV measures views, CPL measures leads, CPI measures installs and CPA measures an agreed action.

Build a controlled test

For CPI Advertising, apply this control to the page's stated scope and evidence window. Use one offer, one landing path, a limited GEO and device scope, consistent conversion tracking and a written stop rule. Review source-level performance before increasing spend. Pause placements that create volume without downstream value, and retain a clean control group so creative, bid and audience changes can be compared.

Use an outcome-normalized score

For CPI Advertising, treat this as a page-specific operating check rather than a universal benchmark. Calculate cost per validated outcome, approval rate, conversion lag, refund or rejection rate, and mature revenue where available. For impression or view pricing, translate spend into the business event that matters. For action pricing, verify the action definition and attribution logic before treating the nominal rate as comparable.

Decision areaQuestion to answerPractical control
BillingWhat exact event creates cost?Document the charge definition and reconcile platform logs.
QualityDoes traffic produce validated outcomes?Use postback or server-side tracking and source reports.
EconomicsWhat is the mature cost per useful result?Include approval, retention, refund and revenue signals.
ScaleCan spend grow without efficiency collapse?Raise budgets gradually and preserve stop thresholds.

Stop and rollback rules

For the CPI Advertising decision, record how this control changes the next test or review. Stop a source when it exceeds the agreed spend cap without enough validated outcomes, when tracking cannot be reconciled, or when downstream quality falls below the business threshold. Roll back to the last stable bid, creative and targeting combination. Do not compensate for weak quality by scaling volume.

Keyword coverage: cpi advertising, cpi marketing.

Search intent and buyer decision

How to use this CPI Advertising: Campaign Setup, Bidding and Optimization page

This URL has one primary job for performance-focused advertisers: decide whether this option fits the buyer's acquisition workflow. Keep this page focused on that buying decision instead of turning it into a generic advertising article. The nearest related FroggyAds page is What Is Cpi Advertising; use that URL when its narrower task is the one you actually need.

The current competitor review for this page records 10 reviewed comparison and competitor pages in the general ads cluster, with 10 fetched successfully. Separately, the page-level entity coverage tracks campaign objective, audience, ad format, budget, bid, conversion tracking, and source quality. We use both as coverage checks, not as copied claims or proof of FroggyAds performance.

StepCommercial General workflowEvidence to retain
1Define the buyer and accepted outcomeKeep the evidence tied to CPI Advertising: Campaign Setup, Bidding and Optimization and the accepted outcome defined for this URL.
2Configure the smallest useful campaign testKeep the evidence tied to CPI Advertising: Campaign Setup, Bidding and Optimization and the accepted outcome defined for this URL.
3Keep, cap or expand only from accepted-outcome evidenceKeep the evidence tied to CPI Advertising: Campaign Setup, Bidding and Optimization and the accepted outcome defined for this URL.

Transparent CPI Advertising: Campaign Setup, Bidding and Optimization decision example

Hypothetical example: if a controlled CPI Advertising: Campaign Setup, Bidding and Optimization test spends USD 225 and records 8 accepted outcomes after the same review window, accepted CPA is USD 225 divided by 8 = USD 28.12. Replace the example inputs with your own economics; this is not a FroggyAds performance claim.

Use FroggyAds as the execution layer only when the page's decision calls for paid traffic. Set the relevant budget, targeting and format controls, verify conversion tracking, keep source-level evidence, and increase spend only when the accepted outcome supports the next step. Create your free FroggyAds account.

Direct answer

CPI Advertising: Campaign Setup, Bidding and Optimization — what matters first

CPI Advertising: Campaign Setup, Bidding and Optimization is most useful when it helps a buyer decide whether this option fits the buyer's acquisition workflow. Define the accepted outcome first, then use targeting, budget and source-level evidence to decide what deserves more spend.