Can a B2B SaaS use Meta's pLTV value optimization for free-trial sign-ups, and what does it require?

TofuAnswer by the Tofu team · checked

Yes, if you get at least 100 Meta-attributed free trials a week and can predict each trial's value within 7 days. Meta's pLTV guide supports the StartTrial event and requires a web Conversions API setup, a validated model, and at least 5 distinct positive values, the highest 3x the lowest. Aimerce reported on August 6, 2026 that pLTV had appeared in Ads Manager, not in every account; Meta hasn't dated it. Use pLTV if you meet all that; use plain value optimization if you have no model; below 100 trials a week, optimize for trials and send CRM stages.

Ask Tofu next

Pick a question and Tofu Ads Agent answers it in chat.

As of November 2026.

Quick answer by situation

If you…UseWhy
Get 100+ Meta-attributed trials a week for 4 weeks and have a tested LTV modelpLTV value optimization on StartTrialMeta optimizes toward the trials you predict are worth most
Have the volume but no modelValue optimization, with an estimated value on StartTrialMeta accepts an "estimated monetary value" and lists no model checks
Get fewer than 100 trials a weekOptimize for trials, and send CRM stages back to MetaYou don't meet pLTV's volume rule
Only know which trials are good after 7 or more daysTrials now, CRM stages laterMeta needs the pLTV value within 7 days of the trial event
Track trials only in your appNot pLTV, for nowMeta says pLTV "does not support app events at this time"

If you have the volume and a model

Meta's guide lists four requirements. You need all of them:

  1. A web Conversions API integration, or one you're ready to build. App events aren't supported.
  2. A pLTV model "validated in some way," such as one you already use for optimization, measurement or budget decisions.
  3. At least 5 distinct predicted values. All must be positive (not binary or negative), and the highest should be at least 3x the lowest.
  4. At least 100 conversion events a week attributed to Meta for each of the last 4 weeks.

The conversion event is where you acquire the user. It happens once per customer and becomes the campaign's optimization event. Meta's own example: "if you can predict pLTV once a user starts a free trial, the free trial is your conversion event." You can tie pLTV to only one event per dataset (pixel). Supported events are Purchase, Subscribe, StartTrial, CompleteRegistration, AddPaymentInfo and custom events.

To set it up, Meta's steps are:

  1. In Events Manager, go to Datasets and select your dataset.
  2. Open the Value and Currency Metrics dropdown and select your trial event. Valid events show Set up pLTV.
  3. Choose how you'll send values. With [Event Name] events sends pLTV with the trial event. After [Event Name] events sends it later. Then select Start integration.
  4. Send the values (next section).
  5. Check health under View status. Once you pass every check, Meta says to allow about 2 weeks for its models to train before your first campaign.
  6. Run a pLTV value optimization campaign for about 7 days to get through the learning phase, and compare it against your current campaign.
  7. Read results in Ads Manager under Columns > Customize columns > Results ROAS, using the pLTV column for your event (Meta's example is Purchase pLTV).

Meta also says: "For campaign setup, budgeting, and scaling guidance, work with your Meta representative."

How to send predicted LTV through the Conversions API

Real time. If your model scores a trial the moment it starts, add predicted_ltv, value and currency to the custom_data of the trial event. The Conversions API is required. The Meta Pixel is optional, and if you use it, send the same parameters.

Delayed, with AppendValue. If your model needs a few days, send the trial event as usual. Then send a separate event named AppendValue with:

  • original_event_data holding the trial's event_name and its event_id and/or order_id, identical to the values on the trial event. Meta uses these to match the two.
  • custom_data with predicted_ltv (a positive number) and currency.
  • user_data with external_id (optional, but Meta recommends it). Hash it the same way on both events.
  • An event_time later than the trial's.

Meta counts the 7 days from when you sent the trial event. If you score trials on day 5, you have at most 2 days left. Sending sooner doesn't help: "sending pLTV values faster does not currently improve performance," Meta says.

If you have the volume but no model

Use standard value optimization instead. Meta's value optimization guide says it "works for all standard and custom events on the Sales objective." You add value and currency to the trial event, the same way in the Pixel and the Conversions API.

The value doesn't have to be revenue. Meta says it can be "estimated monetary value, or other metrics that advertisers identify as key performance indicators," and that its system "values conversions proportional to the value that is passed back." For B2B SaaS, that could be a higher value for trials from companies that match your ideal customer, based on what those trials have been worth to you. Meta's guide lists no model checks for this. The pLTV guide doesn't say which objective pLTV campaigns use.

If you get fewer than 100 trials a week

Then you don't meet pLTV's volume rule, so it isn't an option yet. Optimize for the trial itself, and send Meta later CRM stages (activated, qualified, paid) so it learns which trials matter. Ways to do that:

If you don't have the Conversions API yet, start with setting it up without a developer.

If you only know a trial's worth after 7 days

Many B2B trials only show their value when a buyer converts or talks to sales. pLTV can't wait for that. Meta's window ends 7 days after you send the trial event, and Meta says not to hold back the trial event while you wait for a score. If your model can't give a useful score by then, use CRM stages instead (above).

If your trials start in your app

pLTV is web only for now. Meta says it "does not support app events at this time." For apps, Meta's value optimization guide points to its SDK and Conversions API for App Events guides instead.

What to avoid

  • Delaying the trial event to wait for a score. Meta says to send it as usual and use AppendValue later.
  • Sending yes-or-no values. A 0 or 1 for "qualified" fails the rule of 5 distinct positive values.
  • A new pixel or custom event just to test. Meta advises against it.
  • Comparing ROAS across setups. Meta says ROAS "is comparable only across pLTV campaigns."
  • Pulling final numbers too soon. With delayed pLTV, wait until the campaign ends plus your scoring delay.
  • Assuming you have it. Aimerce says not everyone has the feature, and that a dataset in Core Setup can't use it. Meta's guide doesn't mention Core Setup, and we found no pLTV entry in Meta's 2026 Marketing API changelog.
Sources (4)

Related answers

More on