Google's October 2026 change to Local Services Ads is being described primarily as a billing change.
Starting October 1, certain missed calls and subsequent calls can become chargeable when they meet Google's valid-lead criteria. A missed call during business hours can qualify when the caller remains connected for more than 20 seconds. Google also applies different rules depending on whether the call is answered, reaches voicemail, passes through an IVR, or represents a follow-up interaction with an existing lead. Google Help
The obvious question is:
How many more calls will businesses pay for?
The more interesting question is:
How does an agency determine which customer interaction actually created the billable lead?
That is an attribution problem. And it becomes considerably more complicated when one customer can generate several calls before Google establishes the lead state.
A Lead Is No Longer Just a Call Record
A basic LSA reporting model might look like this:
Call ID
Customer
Timestamp
Campaign
Cost
Lead status
That works reasonably well when the primary question is:
How many calls did the campaign generate?
It becomes less useful when the question changes to:
Which interaction caused Google to classify this customer as a billable lead?
Those are different data problems.
Consider a customer who calls a plumbing company.
10:02 AM → Customer calls
10:02 AM → Business misses call
10:02 AM → Customer disconnects after 12 seconds
10:17 AM → Customer calls again
10:17 AM → Business misses call
10:18 AM → Customer remains connected for 24 seconds
The first call does not satisfy Google's 20-second missed-call condition.
The second interaction potentially does.
If the reporting system simply associates both records with the same customer and labels the customer as "LSA lead," it loses the distinction between the interaction that created the lead and the interaction that did not.
That distinction matters for attribution, billing reconciliation, and campaign optimization.
Google Is Modeling Lead Relationships, Not Just Calls
Google's current documentation says that once a lead has been charged, subsequent calls or messages with the same customer are not charged again when there is interaction within 15 days and the contact comes from the same phone number or email address. After 15 days without activity, a new interaction can create a new lead if it meets the valid-lead criteria.
This introduces a relationship between events.
Instead of:
Call → Charge
the system is closer to:
Customer
│
├── Interaction 1
│
├── Interaction 2
│
├── Interaction 3
│
└── Lead state
│
└── Billing state
The implication for reporting systems is significant.
- A call is an event.
- A lead is a state associated with one or more events.
- A charge is an outcome of that state.
Those concepts shouldn't automatically be stored as the same thing.
The 20-Second Rule Adds Another Dimension
The October change adds temporal conditions to that model.
Google says an unanswered call during business hours can qualify as a valid lead when the customer stays connected for more than 20 seconds. If the business answers the call, the 20-second missed-call rule does not apply; Google evaluates the interaction as an answered call instead. Google Help
That means duration is now part of the attribution context.
A reporting system that records:
missed_call = true
doesn't contain enough information to explain the lead outcome.
At minimum, an agency needs to distinguish something closer to:
answered
missed
connection_duration
business_hours
ad_schedule
ivr_interaction
voicemail
lead_status
charge_status
The exact schema will vary by implementation, but the principle is the same.
The event needs enough context to explain the state transition.
IVR Makes Attribution Even Harder
Google's treatment of automated phone systems makes this particularly interesting.
If a business uses an IVR that asks a caller to press a key to reach the relevant department, the 20-second timer begins after the caller presses the first key. If the caller never presses a key, the missed-call charging condition doesn't apply. Additional key presses in a multi-level menu don't reset the timer.
So this:
Call
↓
IVR
↓
No key pressed
↓
Disconnect
is different from:
Call
↓
IVR
↓
Press 1
↓
20+ seconds
↓
No answer
A call-tracking platform may store both as "missed calls."
Google's lead model does not necessarily treat them the same way. That creates a classic data-integration problem: two systems can describe the same event using different levels of abstraction.
- The telephony system understands routing.
- Google understands lead eligibility.
- The CRM understands customer relationships.
- The billing system understands charges.
Attribution has to connect all four.
Voicemail Creates Another Attribution Edge Case
Google also says that when a call reaches voicemail or an automated greeting during business hours, it can be charged if the caller stays connected for more than 20 seconds, whether or not the caller leaves a message.
This creates an important distinction between:
Customer leaves voicemail
and:
Customer reaches voicemail
Customer stays connected >20 seconds
Customer leaves no message
A CRM may treat only the first as an actionable lead.
Google's lead system can treat both differently from a simple "missed call" classification.
That means agencies should not assume the CRM's definition of a lead is interchangeable with Google's definition of a billable lead.
Attribution Breaks When Systems Use Different Customer Identifiers
The problem becomes harder when the same customer contacts the business through different identifiers.
Suppose the first call comes from:
(555) 111-2222
The second call comes from:
(555) 111-3333
The customer then sends a message using an email address.
The CRM may recognize all three interactions as one customer. A basic call-reporting system may not.
Google's current documentation says its 15-day follow-up treatment relies on the same phone number or email address. This means identity resolution matters.
An agency that wants to reconcile Google lead data against CRM outcomes cannot simply join everything on customer name.It needs a deliberate identity model.
For example:
Customer
│
├── Phone identity
├── Email identity
├── CRM contact ID
├── Call IDs
└── Lead IDs
Without that layer, the same customer can appear to be multiple leads. Or multiple interactions can be incorrectly attributed to one lead. Both errors distort cost-per-lead calculations.
Business Hours Are Now Attribution Data
There is another dependency that can easily be missed.
Google defines business hours using the hours configured in the Google Business Profile. If an ad schedule differs from those hours, an unanswered call is chargeable under the missed-call rule only when it occurs during both the active ad schedule and business hours. Google Help
That means a reporting system attempting to explain a charge may need more than:
campaign_id
call_timestamp
call_duration
It may also need to understand the availability state at that exact time.
Conceptually:
Call
↓
Campaign active?
↓
Within business hours?
↓
Answered?
↓
IVR state?
↓
Duration?
↓
Existing lead?
↓
Valid lead?
↓
Chargeable?
The attribution system is therefore becoming a rules engine. Not necessarily a complicated one, but a state-dependent one.
The Migration Makes the Problem More Important
This change is happening while Google is moving Local Services Ads into Performance Max campaigns with pay-per-lead goals.
Google says the migrated campaigns retain the pay-per-lead model, while campaign management moves into Google Ads. Google also says performance reports from the previous LSA environment will not transition into Google Ads. Google Help
That creates two separate data challenges. The first is historical continuity. The second is lead attribution.
An agency now has to answer:
Can I compare historical LSA leads with post-migration leads?
But also:
Can I explain why a post-migration interaction became a billable lead?
Those are different questions.
A unified dashboard doesn't automatically solve either one.
The Reporting Model Should Separate Events, Leads, and Revenue
This is where I think agencies need to change their internal data models.
Instead of treating:
Lead = Call = Conversion = Customer
model the lifecycle separately:
Interaction
↓
Lead
↓
Qualified opportunity
↓
Booked job
↓
Revenue
- A single lead can contain multiple interactions.
- A single customer can generate multiple leads over time.
- A single interaction may not become a billable lead.
- And a billable lead may never become revenue.
Those distinctions aren't academic.
If an agency optimizes toward the wrong layer, it can make the campaign look healthier while the business becomes less profitable.
For example:
100 interactions
↓
70 billable leads
↓
45 qualified opportunities
↓
28 booked jobs
↓
18 completed jobs
↓
$42,000 revenue
If the agency only reports the first two numbers, it knows how efficiently Google generated leads. It doesn't know how efficiently those leads generated business.
What Agencies Should Add to Their Attribution Model
After the October update, I would want an LSA reporting system to distinguish at least four layers.
Interaction layer
What actually happened?
- Call
- Message
- Booking
- Voicemail
- Follow-up interaction
Lead layer
Did Google classify the interaction as a valid lead?
- Valid
- Not charged
- Charged
- Credited
- Pending/review
Google's automated lead-credit system can reassess charged leads over time, with credits potentially appearing later if a lead is determined to be low quality or invalid. Google Help
Customer layer
Who did the interaction belong to?
- Phone
- CRM contact
- Existing customer
- New customer
Revenue layer
What happened after the lead?
- Qualified
- Appointment
- Estimate
- Won
- Lost
- Revenue
The important part is not creating four more dashboards.
It's maintaining the relationships between those layers.
The Real Attribution Problem
The October LSA update doesn't necessarily make attribution impossible.
It makes simplistic attribution models less reliable.
A system that says:
Campaign → 100 calls → 100 leads
cannot explain the increasingly nuanced path between an interaction and a billable lead.
A better model looks like:
Campaign
↓
Interaction
↓
Customer identity
↓
Lead state
↓
Billing state
↓
CRM outcome
↓
Revenue
That model also gives agencies something much more valuable than another monthly lead report. It gives them an audit trail.
If a client asks why lead costs increased, the agency can investigate whether the change came from more missed calls, longer unanswered calls, different IVR behavior, new customers, [follow-up interactions,](https://Can a follow-up call become a billable LSA lead if the original call was not charged?) lead credits, or a change in downstream conversion rates.
Without that structure, all those events collapse into one number: Cost per lead. And that number may be accurate while still failing to explain what actually happened.
The Bigger Change
Google's October 2026 LSA update is easy to summarize as a rule about missed and follow-up calls. But the more important technical change is what it does to the definition of a lead.
A lead is no longer something an attribution system can safely model as a single phone event. It can be the result of a sequence of interactions, timing conditions, routing states, customer identity, and previous lead activity.
That makes LSA reporting increasingly similar to an event-driven system.
The question is no longer simply "Which campaign generated this call?"
The better question is:
"Which interaction caused this customer to enter a billable lead state, and what happened after that?"
For agencies building reporting pipelines around Google's new LSA-to-Google Ads environment, that distinction may matter more than the dashboard itself.