Reply link tracking and attribution
You can now track what happens after someone clicks a shortened link shared in a BrandBastion reply, making it easier to connect social engagement to traffic and conversions. New dynamic UTM tracking automatically captures the context behind each click, while click reporting shows which sources, campaigns, and content are performing best.
Note: This is only available for links shortened through BrandBastion’s link shortening integration. Available for Agent+ plans.
Dynamic UTMs for shortened links
Automatically add tracking parameters to shortened links in replies based on conversation context, including platform, message type, campaign name, and campaign ID. This gives you consistent attribution without manually creating tracked URLs for every reply.
Top clicked UTMs
See which UTM values are generating the most clicks from links shared in replies. Switch between dimensions like Source, Campaign, and Content to understand what is driving traffic and identify your top-performing reply strategies.
Use cases:
Measure which platforms, reply types, and campaigns drive the most link clicks and conversions
Connect reply engagement back to specific campaigns without manually building tracked URLs
Compare performance across UTM sources, campaigns, and content variants
Sentiment classification improvements
BrandBastion is implementing an improvement to our sentiment classification models. You may notice more accurate sentiment results, particularly for comments where context, intention, or sarcasm matter. We can also make targeted adjustments for recurring scenarios that a brand interprets differently.
For example, comments that objectify a person featured in content may previously have been marked neutral or negative, depending on their wording and context. They can now be classified as positive when they reflect welcome fan behaviour and the brand considers that behaviour positive.
Use cases:
Improve accuracy for comments where tone and intent matter more than individual words.
Adapt sentiment for recurring, brand-specific scenarios without changing the overall sentiment framework.
Keep reporting aligned with how your team interprets key types of conversation.
Monitor TikTok ad performance in BrandBastion
Review TikTok campaign, ad group, and ad performance alongside BrandBastion’s sentiment, comment, and moderation insights. You can now understand how your TikTok ads are performing and how audiences are responding without switching between TikTok Ads Manager and BrandBastion.
Use cases:
Compare spend, reach, clicks, and conversions with sentiment and comment volume.
Identify campaigns or ads attracting high engagement but increased negative feedback.
Review performance and moderation outcomes across campaigns, ad groups, and individual ads.
Astra analysis in ad performance
Astra is now available directly inside Ad Performance, so you can analyze the exact table you are looking at and instantly see what is driving results, what is slipping, and what to do next.
Use cases:
Identify which campaigns are worth scaling by comparing ROAS and CTR alongside CM score and comment signals in the same view.
Spot wasted spend fast by flagging high-spend items with weak returns, then linking performance drag to low response rate, rising negative sentiment, or low moderation rate.
Turn analysis into action by pulling a short list of next steps, like where to tighten moderation, adjust reply coverage, or replicate top-performer patterns across similar ad sets.
Filter to report by triggered workflow
You can now filter conversations by the workflows triggered on them, making it easier to investigate how specific conversations were handled and report on workflow activity. Workflows are grouped by type, including replies, likes, hides, alerts, tags, assignments and case creation.
Use cases:
Find conversations handled by a specific workflow.
Report how many conversations were liked, replied to, hidden or otherwise actioned.
Measure and compare the activity of individual workflows or workflow types.
PR, collaboration & spam tag improvements
The PR & Collaboration tag now better identifies both inbound partnership interest and comments or DMs suggesting potential collaborators for your brand. Spam tagging has also been extended to catch additional scam DMs, such as fake Meta account warnings.
Use cases:
Surface creators, users, and community members who want to work with your brand.
Identify suggestions to collaborate with a specific creator, brand, or public figure.
Filter out clear scam DMs, such as messages falsely claiming Meta will delete your account.
Full reply composer tools in batch mode
Batch mode now includes link shortening, attachments, products and other supported reply composer actions. You can also choose which connected account or identity the reply should be sent from directly within the batch approval workflow.
Use cases:
Add tracked or shortened links while approving replies in bulk.
Attach images, products or other supported content without leaving batch mode.
Select the correct replying identity for each conversation before approval.
Improved AI reply experience
We’ve made the AI reply experience clearer and easier to use across Agent+ and Engage+. The updates give agents more visibility into how replies are generated and make it easier to understand, find, and switch between reply prompts.
See how the AI generates replies: Follow each step while Astra reads the conversation, analyzes the related post when relevant, checks your brand knowledge, matches your tone of voice, and drafts the reply.
See which prompt was used: After generation, Astra displays the name of the prompt used and keeps it visible alongside the reply.
Clearer prompt controls: Updated tooltips make it clear which prompt will be used to generate a reply and how to select a different one.
Search AI prompts: Search prompts by name directly from the prompt selector to quickly find the right one.
Use cases:
Understand what the AI is considering when generating a reply.
Confirm which prompt was used and easily switch between different response strategies.
Find the right prompt faster when your team has a larger prompt library.
Contextual loading updates for Astra
Astra now shows contextual progress updates while it analyzes your data, so you can see what it is working on.
Astra in Analyze: See progress as Astra reviews the selected conversations and date range, identifies recurring topics, analyzes sentiment, and prepares its findings.
Astra in Posts: Loading updates reflect the selected posts and date range, including steps like reviewing engagement, analyzing comment sentiment, and identifying the posts driving results.
Astra in Campaigns: Astra shows progress as it reviews selected campaigns, compares delivery and engagement metrics, and connects campaign performance with comment sentiment.
Use cases:
Understand what Astra is analyzing while a request is still running.
See how your selected date range, posts, campaigns, or conversations are being used.
Where to find it: Open Astra from Analyze, Posts, or Campaigns. The relevant loading updates appear automatically while Astra is generating its response.
More flexible scheduled reporting
Scheduled reports can now follow your own reporting calendar. Choose when your reporting day, week, month, or quarter starts, and BrandBastion will automatically adjust the reporting period included in each scheduled report.
Use cases:
Align weekly reports with a business week that starts on Saturday, Sunday, or another day.
Match monthly or quarterly reporting with internal or fiscal reporting periods.
Set a specific start time for daily reporting so data aligns with your team's operating hours.
Easier tracking for replies sent outside BrandBastion
When a conversation is marked as replied outside the platform, BrandBastion now automatically records who replied and when. You can optionally add the reply text, edit the attribution or timestamp, and undo the action if it was marked by mistake. These responses are also included in the attributed agent’s reply and performance metrics.
Use cases:
Track replies handled directly on a social network, email, or another platform.
Keep agent performance reporting accurate when responses happen outside BrandBastion.
Quickly correct conversations that were accidentally marked as replied.
