Tips for how to manage the impact of marketing industry and regulatory shifts on your website campaigns

We continue to adapt our systems to help you reach your customers, improve business outcomes and deliver conversions, while honouring consumer privacy. Learn how marketing industry and regulatory shifts may affect your ads and reporting.

As you build your website campaigns across Meta technologies, use these tips to help you manage the performance impacts from marketing industry and regulatory shifts:

• Implement the Conversions API: We recommend that you implement the Conversions API and send your top eight events configured in the web event configuration tool through the Conversions API. This action can help you maximise benefit from our ongoing updates. Learn how you can increase performance with the Conversions API and simplify integrations with the Conversions API Gateway.
  
• Exit the learning phase: Learn about the learning phase and tips for how to exit it.
  

  • Aim for at least 50 conversions per week: Ad sets usually exit the learning phase after around 50 optimisation events since its last significant edit. You can also consider increasing budgets and/or prioritising your higher-frequency, lower-funnel web events and the events that represent the largest, most diverse audiences.
    
  • Consolidate and reduce the number of ad sets in the learning phase: This may allow for an increased budget and audience size, and can help meet the 50 conversions-per-week threshold that's necessary to exit the learning phase. You can also consider reducing the frequency of manual edits by making fewer, more strategic changes to prevent your ads from returning to the learning phase. Learn how to combine ad sets and campaigns.
    
• Analyse reporting at the campaign level: While ad set and ad-level assessment is possible, some estimated conversions may be reported after a delay. As a result, they may not reflect true conversion time for modelled conversions. We also recommend that you consider waiting a minimum of 72 hours or the full length of the optimisation window selected before evaluating performance.