TL;DR: Google eliminated all attribution models except two: Last-Click and Data-Driven Attribution (DDA). DDA delivers 6-30% more conversions and 20-30% cost reduction for qualifying accounts. But it requires approximately 3,000 ad interactions and 300 conversions in 30 days to function properly. If your account falls below these thresholds, DDA may actually hurt performance. This guide explains how each model works, when to use which, and how your choice directly impacts Smart Bidding decisions.
Attribution is not a theoretical concept. It directly controls where Google's Smart Bidding sends your budget. Choose wrong, and you optimize toward the wrong outcomes.
What Attribution Models Actually Do
Attribution models answer one question: when a customer interacts with your ads multiple times before converting, which interaction gets credit?
Why This Matters for Service Businesses
Consider a typical service business customer journey:
- Day 1: Searches "best plumbers near me," clicks your ad, browses your site, leaves
- Day 5: Searches "plumber reviews [your city]," clicks your ad again, reads reviews, leaves
- Day 8: Searches "[your company name]," clicks your branded ad, calls you, becomes a customer
Three ad clicks. One conversion. Which click gets credit?
The answer directly affects Smart Bidding:
- If only the last click (branded search) gets credit, Smart Bidding sends more budget to branded campaigns
- If credit is distributed across all three clicks, Smart Bidding recognizes the value of "best plumbers near me" and "plumber reviews" keywords too
The wrong model can systematically starve your top-of-funnel keywords of budget.
The Two Remaining Models
In September 2023, Google eliminated Linear, Time Decay, Position-Based, and First-Click attribution models. Only two remain.
Model 1: Last-Click Attribution
How it works: 100% of the conversion credit goes to the last ad click before the conversion.
In the plumber example:
| Interaction | Credit |
|---|---|
| Day 1: "best plumbers near me" | 0% |
| Day 5: "plumber reviews [city]" | 0% |
| Day 8: "[company name]" (converted) | 100% |
Strengths:
- Simple, easy to understand
- Deterministic (no modeling or estimation)
- Works with any data volume
- Best for single-interaction conversion paths
Weaknesses:
- Ignores all touchpoints before the final click
- Over-credits branded and low-funnel keywords
- Under-credits awareness and consideration keywords
- Can cause budget misallocation in multi-touch journeys
Model 2: Data-Driven Attribution (DDA)
How it works: Google's machine learning analyzes your account's actual conversion paths and distributes credit based on each touchpoint's incremental contribution to the conversion.
In the plumber example (hypothetical DDA distribution):
| Interaction | Credit |
|---|---|
| Day 1: "best plumbers near me" | 40% |
| Day 5: "plumber reviews [city]" | 25% |
| Day 8: "[company name]" (converted) | 35% |
DDA recognized that the initial "best plumbers" search was actually the most influential interaction because without it, the customer would never have discovered the business.
Strengths:
- Reflects actual customer behavior in your specific account
- Recognizes the value of upper-funnel keywords
- Better budget allocation across the full funnel
- Continuously learns and adapts
Weaknesses:
- Requires significant data volume to work properly
- Black box (you cannot see exactly how credit is distributed)
- Can be unreliable for small accounts
- Affected by iOS privacy changes
DDA Data Requirements
DDA is the default model for new conversion actions. But "default" does not mean "appropriate for every account."
Minimum Thresholds
| Requirement | Threshold | Time Period |
|---|---|---|
| Ad interactions (clicks) | ~3,000 | 30 days |
| Conversions | ~300 | 30 days |
If your account falls below these thresholds, DDA does not have enough data to model attribution accurately. It reverts to a simplified model internally, which may be less accurate than simple Last-Click.
Threshold Reality Check for Service Businesses
Let us run the numbers for a typical service business:
| Monthly Spend | Average CPC | Monthly Clicks | Monthly Conversions (7.5% CVR) | DDA Qualified? |
|---|---|---|---|---|
| $1,000 | $5.26 | ~190 | ~14 | No |
| $2,500 | $5.26 | ~475 | ~36 | No |
| $5,000 | $5.26 | ~950 | ~71 | No |
| $10,000 | $5.26 | ~1,900 | ~143 | No |
| $15,000 | $5.26 | ~2,850 | ~214 | Borderline |
| $20,000 | $5.26 | ~3,800 | ~285 | Close |
| $25,000 | $5.26 | ~4,750 | ~356 | Yes |
The uncomfortable truth: Most service businesses spending under $20,000/month on Google Ads do not generate enough data for DDA to work properly.
Using the overall average CPC of $5.26 and average conversion rate of 7.52%, you need roughly $20,000-$25,000/month in spend to reliably hit DDA thresholds.
Higher-CPC industries (legal at $8.58, dental at $7.85) need even more spend. Lower-CPC industries might qualify at lower spend levels.
How to Check If Your Account Qualifies
- Go to Google Ads > Goals > Conversions
- Click on your conversion action
- Under Attribution model, check if DDA is available
- If Google shows a warning about insufficient data, your account does not qualify
Alternatively, check your Conversion Paths report:
- Tools & Settings > Attribution > Conversion Paths
- If most paths show only 1 touchpoint, DDA provides minimal benefit anyway
DDA Performance Impact
For accounts that DO meet the data thresholds, DDA delivers significant improvements.
The Numbers
| Metric | DDA Improvement | Source |
|---|---|---|
| Conversion increase | 6-30% | Multiple studies |
| Cost-per-conversion reduction | 20-30% | Multiple studies |
| Cross-device attribution | Improved | |
| Upper-funnel keyword recognition | Significantly better | Industry consensus |
Why DDA Outperforms Last-Click
Scenario: Law firm spending $15,000/month
Under Last-Click:
| Campaign | Conversions (Last-Click) | Budget Allocated |
|---|---|---|
| Branded searches | 30 | $5,000 (33%) |
| "best lawyer near me" | 8 | $3,000 (20%) |
| "lawyer reviews [city]" | 5 | $2,000 (13%) |
| "personal injury attorney" | 7 | $5,000 (33%) |
Smart Bidding sees branded campaigns converting best and pushes budget there.
Under DDA:
| Campaign | Conversions (DDA) | Budget Allocated |
|---|---|---|
| Branded searches | 18 (partial credit) | $3,000 (20%) |
| "best lawyer near me" | 15 (gets discovery credit) | $5,000 (33%) |
| "lawyer reviews [city]" | 10 (gets consideration credit) | $3,000 (20%) |
| "personal injury attorney" | 12 (gets intent credit) | $4,000 (27%) |
DDA redistributes credit to the keywords that actually initiate customer journeys. Smart Bidding then allocates budget more effectively across the full funnel.
Result: More total conversions at lower cost because budget flows to truly valuable keywords instead of just the last-touch branded terms.
When to Use Each Model
Use Last-Click When:
| Scenario | Why Last-Click Works |
|---|---|
| Monthly spend under $15,000 | Not enough data for DDA |
| Fewer than 100 monthly conversions | DDA cannot model accurately |
| Single-interaction conversions | Most customers click once and convert |
| Simple conversion paths | No multi-touch journeys to model |
| Emergency services (plumber, locksmith, towing) | Customers search once and call immediately |
| New accounts (first 3-6 months) | Building data for eventual DDA switch |
Emergency services are a key example. Someone with a burst pipe searches "emergency plumber near me," clicks the first credible ad, and calls immediately. There is one touchpoint. DDA adds no value. Last-Click is simpler and equally accurate.
Use DDA When:
| Scenario | Why DDA Works |
|---|---|
| Monthly spend $20,000+ | Sufficient data volume |
| 300+ monthly conversions | DDA can model accurately |
| Multi-touch conversion paths | Customers research before buying |
| Professional services (legal, financial, consulting) | Long consideration cycles |
| High-value services (remodeling, construction) | Multiple research touchpoints |
| Running both branded and non-branded campaigns | DDA properly credits each |
| Using Smart Bidding strategies | DDA feeds better signals |
The Transition Path
For growing accounts, plan the transition:
Phase 1: Start with Last-Click (Months 1-6)
- Build data, optimize campaigns
- Monitor conversion path lengths in Attribution reports
- Track whether customers interact with multiple keywords before converting
Phase 2: Evaluate DDA readiness (Month 6)
- Check if you meet the 3,000 interactions / 300 conversions thresholds
- Review Conversion Paths report for multi-touch patterns
- If most conversions are single-touch, Last-Click may remain optimal
Phase 3: Switch to DDA (when thresholds met)
- Switch the attribution model on your primary conversion action
- Expect a learning phase for Smart Bidding (7-14 days)
- Monitor closely: conversion volume should stabilize or improve
- Compare 30-day pre/post metrics
Phase 4: Monitor and validate (ongoing)
- Check Model Comparison report monthly
- Compare Last-Click vs DDA conversion counts
- If DDA consistently shows more conversions for non-branded campaigns, it is working correctly
How Attribution Affects Smart Bidding
Your attribution model is not just a reporting preference. It fundamentally changes how Smart Bidding (Target CPA, Target ROAS, Maximize Conversions) allocates your budget.
The Mechanism
Smart Bidding uses conversion data to decide how much to bid for each auction. It asks: "Based on historical data, how likely is this specific search query, at this time, on this device, in this location, to lead to a conversion?"
The attribution model determines what "conversion data" Smart Bidding has to work with.
Under Last-Click: Smart Bidding sees that "emergency plumber" got 5 conversions and "[company name]" got 15 conversions. It bids aggressively on branded terms and conservatively on non-branded.
Under DDA: Smart Bidding sees that "emergency plumber" actually contributed to 12 conversions (including partial credit for multi-touch paths). It bids more aggressively on non-branded discovery keywords.
The Budget Allocation Cascade
| Attribution Model | Budget Flow | Smart Bidding Behavior |
|---|---|---|
| Last-Click | Concentrates on bottom-funnel | Bids high on branded, low on discovery |
| DDA | Distributes across funnel | Balances bids across all touchpoints |
Practical Implications
If you switch from Last-Click to DDA, expect:
- Branded campaign conversions to decrease (credit redistributed)
- Non-branded campaign conversions to increase (credit gained)
- Smart Bidding to redistribute budget (more to non-branded)
- A learning phase of 7-14 days (Smart Bidding recalibrates)
- Net total conversions to increase (typically 6-30%)
Warning: If you switch models when your account does not meet DDA thresholds, the opposite may happen. Smart Bidding gets confused by unreliable DDA signals and performance degrades.
iOS Privacy Impact on Attribution
Apple's App Tracking Transparency (ATT) framework, introduced with iOS 14.5, significantly impacts attribution accuracy.
The Data Loss
| Platform | Impact on Observable Data |
|---|---|
| iOS (Safari) | 18-32% reduction in observable conversion data |
| iOS (Apps) | Significant tracking limitations |
| Android | Minimal impact (for now) |
| Desktop | Moderate (cookie restrictions increasing) |
What This Means for Attribution
When iOS users click your ad and convert, the conversion may not be visible to Google Ads if:
- The user opts out of tracking (most do)
- Safari blocks third-party cookies
- The conversion happens on a different device
DDA is more affected than Last-Click because DDA needs to observe full conversion paths across multiple interactions. If iOS blocks visibility of intermediate touchpoints, DDA's model becomes less accurate.
Mitigation Strategies
| Strategy | Effectiveness | Effort |
|---|---|---|
| Enhanced Conversions | High (recovers email/phone match) | Low |
| Server-side tracking | Very High (13-27% accuracy improvement) | Medium-High |
| First-party data strategy | High (own your customer data) | Medium |
| Consent mode v2 | Medium (models consented users to estimate total) | Low |
The combined approach: Enhanced Conversions + server-side tracking recovers most of the iOS data loss. For accounts spending $5,000+/month, this combination is essential.
Related reading: For the complete server-side tracking and Enhanced Conversions setup, see Conversion Tracking for Service Businesses: Calls, Forms, and Offline Revenue.
Cross-Device Tracking
Service business customers frequently start their research on mobile and convert on desktop (or vice versa).
The Cross-Device Challenge
| Journey | Attribution Challenge |
|---|---|
| Mobile search > Desktop call | Mobile click gets no conversion credit without cross-device tracking |
| Desktop research > Mobile call | Desktop clicks undervalued |
| Tablet browse > Phone call | Tablet interactions invisible |
How Google Handles Cross-Device
Google uses signed-in user data (Google account activity) to connect interactions across devices. If the same Google account clicks an ad on mobile and later converts on desktop, Google can attribute the conversion.
Limitations:
- Only works for signed-in Google users
- Privacy regulations may limit data availability
- Not all cross-device journeys are captured
DDA handles cross-device better than Last-Click because it can assign partial credit to the mobile click even when the conversion happens on desktop. Last-Click only credits the desktop interaction.
Viewing Cross-Device Data
- Go to Tools & Settings > Attribution > Cross-Device Activity
- View device path combinations
- Identify which device starts the journey vs. which completes it
For most service businesses, you will see:
- Mobile starts 60-70% of conversion paths
- Desktop/phone call completes 40-50% of conversions
- Cross-device paths are typically 2-3 interactions
The Model Comparison Report
Google Ads provides a direct comparison of how different models would attribute your conversions.
How to Access
Tools & Settings > Attribution > Model Comparison
How to Read It
The report shows your conversion data attributed under both Last-Click and DDA. Compare:
| Campaign | Last-Click Conversions | DDA Conversions | Difference |
|---|---|---|---|
| Branded | 50 | 35 | -30% (DDA gives less credit) |
| Non-branded generic | 20 | 30 | +50% (DDA gives more credit) |
| Non-branded service | 15 | 25 | +67% (DDA gives more credit) |
| Total | 85 | 90 | +6% |
How to interpret:
- Campaigns gaining credit under DDA are contributing more value than Last-Click reveals
- Campaigns losing credit under DDA are getting over-credited by Last-Click
- If DDA shows significantly more total conversions, switching models will improve Smart Bidding
Red Flags in the Comparison
| Pattern | What It Means | Action |
|---|---|---|
| DDA shows 20%+ fewer branded conversions | Branded campaigns are over-credited under Last-Click | Consider switching to DDA |
| DDA shows nearly identical numbers | Most conversions are single-touch | Last-Click is fine; DDA adds little value |
| DDA shows wildly different numbers | Multi-touch journeys are common | DDA is strongly recommended |
| DDA shows fewer total conversions | Possible data quality issue | Audit conversion tracking before switching |
Attribution Strategy by Business Type
Emergency Services (Plumbing, HVAC, Locksmith, Towing)
Recommended model: Last-Click
Why: Conversion paths are almost always single-touch. Customer has an emergency, searches once, clicks, calls. Multi-touch modeling adds complexity without value.
Conversion path length: 1 interaction (90%+ of conversions)
Professional Services (Law, Accounting, Financial Advisory)
Recommended model: DDA (if thresholds met) or Last-Click (if not)
Why: Customers research extensively. A personal injury client may search 5-10 times across 2-3 weeks before calling a lawyer. DDA properly credits the initial discovery searches.
Conversion path length: 3-7 interactions typical
Home Improvement (Remodeling, Landscaping, Roofing)
Recommended model: DDA (if thresholds met) or Last-Click (if not)
Why: High-value projects require research. Homeowners compare multiple providers, read reviews, visit websites repeatedly. DDA captures this multi-touch journey.
Conversion path length: 2-5 interactions typical
Health & Wellness (Dentists, Chiropractors, Med Spas)
Recommended model: Last-Click (most accounts) or DDA (large practices)
Why: Most dental patients search once and book. Larger practices with higher budgets may see multi-touch paths for elective procedures.
Conversion path length: 1-2 interactions typical (routine), 3-5 for elective
Recurring Services (Cleaning, Pest Control, Lawn Care)
Recommended model: Last-Click
Why: Relatively simple conversion paths. Customer searches, clicks, books. The first service is low-risk and low-commitment, reducing the need for research.
Conversion path length: 1-2 interactions typical
Decision Framework: Which Model to Choose
Use this framework to make the decision:
Step 1: Check Data Volume
Do you have 3,000+ clicks and 300+ conversions per month?
- Yes: DDA is an option. Proceed to Step 2.
- No: Use Last-Click. Revisit when you reach thresholds.
Step 2: Check Conversion Path Length
Go to Attribution > Conversion Paths. Are most paths multi-touch (2+ interactions)?
- Yes (30%+ of paths are multi-touch): DDA adds significant value. Proceed to Step 3.
- No (80%+ are single-touch): Last-Click is fine. DDA adds minimal value.
Step 3: Check the Model Comparison
Go to Attribution > Model Comparison. Does DDA redistribute credit meaningfully?
- Yes (non-branded campaigns gain 20%+ credit): Switch to DDA.
- No (numbers are nearly identical): Last-Click is adequate.
Step 4: Monitor After Switching
After switching to DDA, did total conversions increase within 30 days?
- Yes: DDA is working. Keep it.
- No (conversions decreased): Either data volume is insufficient or conversion paths are too simple. Revert to Last-Click.
Common Attribution Mistakes
Mistake 1: Using DDA with Insufficient Data
Impact: DDA's model is unreliable, Smart Bidding gets confused signals, performance degrades.
Fix: Check the thresholds. 3,000 interactions + 300 conversions in 30 days. If you are under, use Last-Click.
Mistake 2: Ignoring Attribution Entirely
Impact: You accept whatever default Google sets (DDA), regardless of whether your account qualifies.
Fix: Actively choose your attribution model. Check data volume. Review the Model Comparison report quarterly.
Mistake 3: Switching Models Without a Monitoring Plan
Impact: You switch to DDA and do not notice that performance degraded because you did not set a baseline.
Fix: Before switching: document current CPA, conversion volume, and ROAS for 30 days. After switching: compare against baseline at 14, 30, and 60 days.
Mistake 4: Over-Crediting Branded Campaigns
Impact: Under Last-Click, branded campaigns look like conversion machines. You shift budget there, starving the non-branded campaigns that actually create demand.
Fix: Check the Model Comparison report. If branded campaigns lose significant credit under DDA, your non-branded campaigns are undervalued.
Mistake 5: Not Accounting for iOS Data Loss
Impact: Your attribution model operates on incomplete data. DDA's model is distorted by the 18-32% of conversions it cannot see from iOS users.
Fix: Implement Enhanced Conversions and server-side tracking to recover lost data. The more complete your data, the more accurate any attribution model becomes.
Mistake 6: Treating Attribution as Set-and-Forget
Impact: Your business grows, data volume changes, customer behavior shifts. The model that was right 6 months ago may not be right today.
Fix: Quarterly review of the Model Comparison report. Quarterly check of data volume vs DDA thresholds. Annual reassessment of conversion path patterns.
Attribution Reporting: What to Monitor
Beyond choosing the right model, you need to actively monitor attribution data to make informed decisions.
Reports to Review Monthly
1. Conversion Paths Report
- Location: Tools & Settings > Attribution > Conversion Paths
- What to look for: Average path length, most common keyword sequences, device transitions
- Action: If average path length is growing, your audience is taking longer to decide. Consider adjusting conversion windows.
2. Model Comparison Report
- Location: Tools & Settings > Attribution > Model Comparison
- What to look for: Campaigns where DDA and Last-Click differ by more than 20%
- Action: If non-branded campaigns gain significant credit under DDA, they are undervalued under Last-Click.
3. Time to Conversion Report
- Location: Tools & Settings > Attribution > Conversion Paths > Time Lag
- What to look for: What percentage of conversions happen same-day vs. 7+ days later
- Action: If 40%+ of conversions happen after 7 days, your conversion window needs to be at least 30 days.
4. Assisted Conversions Report
- Location: Google Analytics 4 > Advertising > Attribution > Conversion Paths
- What to look for: Keywords with high assist counts but low last-click conversions
- Action: These are your hidden value keywords. DDA would properly credit them; Last-Click does not.
Creating an Attribution Dashboard
Track these metrics monthly:
| Metric | Data Source | Purpose |
|---|---|---|
| Average conversion path length | Conversion Paths report | Track journey complexity over time |
| % single-touch vs multi-touch | Conversion Paths report | Validate model choice |
| DDA vs Last-Click variance by campaign | Model Comparison | Identify undervalued campaigns |
| Time to conversion (median) | Time Lag report | Set appropriate conversion windows |
| Cross-device conversion % | Cross-Device report | Understand device transitions |
| iOS vs Android conversion rates | Device segmentation | Monitor privacy impact |
The Bigger Picture: Attribution in Your Measurement Stack
Attribution is one component of a complete measurement strategy. It works best when combined with:
1. Complete Conversion Tracking
Attribution can only credit what it can see. If phone calls are untracked, no attribution model can properly value the keywords that generate calls.
Related reading: Conversion Tracking for Service Businesses: Calls, Forms, and Offline Revenue
2. Offline Conversion Data
For service businesses, the most valuable conversion (a signed contract or completed job) happens offline. Uploading this data back to Google Ads gives both Last-Click and DDA more accurate signals to work with.
3. Server-Side Tracking
The 13-27% accuracy improvement from server-side tracking directly improves attribution accuracy. More data points mean better models.
4. Primary vs Secondary Conversion Configuration
Attribution only applies to your Primary conversion actions. If you have the wrong actions set as Primary, even perfect attribution will optimize toward the wrong outcomes.
Key Takeaways
Attribution Models in 5 Points
- Only 2 models remain — Last-Click (simple, works everywhere) and Data-Driven Attribution (ML-powered, needs data)
- DDA requires ~3,000 clicks + 300 conversions/month — Most service businesses under $20K/month spend do not qualify
- DDA delivers 6-30% more conversions for qualifying accounts — With 20-30% cost reduction
- Your model directly controls Smart Bidding — Wrong model = wrong budget allocation
- iOS privacy reduces observable data 18-32% — Enhanced Conversions + server-side tracking are essential mitigations
Decision Summary
| Your Account | Recommended Model |
|---|---|
| Under $15K/month spend | Last-Click |
| $15-25K/month, simple paths | Last-Click |
| $15-25K/month, multi-touch paths | Test DDA, monitor closely |
| $25K+/month, multi-touch paths | DDA |
| Emergency services (any budget) | Last-Click |
| Professional services ($20K+/month) | DDA |
Priority Actions
| Action | Impact | Effort |
|---|---|---|
| Check if your account meets DDA thresholds | Critical | Low |
| Review Model Comparison report | High | Low |
| Review Conversion Paths for multi-touch patterns | High | Low |
| Enable Enhanced Conversions (for any model) | High | Low |
| Implement server-side tracking ($5K+ spend) | Very High | High |
| Set quarterly attribution model review | Medium | Low |
| Document baseline metrics before any model switch | High | Low |
What Doesn't Matter
- Debating attribution models when you have fewer than 100 monthly conversions (use Last-Click)
- Chasing "perfect" attribution (it does not exist; all models are approximations)
- Switching models frequently (each switch triggers a learning phase)
What Actually Matters
- Having enough data for your chosen model to work properly
- Matching attribution model to your actual customer journey patterns
- Feeding complete conversion data (including calls and offline) into whatever model you use
- Monitoring performance after any model change
- Quarterly reassessment as your account grows
This guide is part of the Google Ads Efficiency Playbook 2026 series. Data sourced from Google Ads documentation, Growth Minded Marketing, Fibbler, Optmyzr (Frederick Vallaeys), and multiple independent performance studies.