Playbook: Reading Social Performance Without Drowning in Metrics

The Problem with Raw Volume

When you monitor a brand with decent traction, the absolute volume of data becomes the primary bottleneck for your marketing operations team. In a recent 30-day window, a standard mid-market consumer brand generated exactly 14,310 scored mentions across its active tracking monitors.

If your social team attempts to read 14,310 mentions sequentially, they will fail. If they try to rank them strictly by raw follower count, they will miss highly engaged niche conversations that actually drive product adoption. The traditional industry approach is to build complex Boolean queries to filter out the noise, but this often results in severe false negatives. You end up missing the actual signal because you over-tuned the filters.

To solve this, Sentia relies on a specific methodology we call "bands not points", powered by the mention_priority_score (MPS). This playbook walks you through how to configure your Sentia workspace to turn an overwhelming feed into a finite, highly actionable checklist of daily decisions.

Understanding the MPS Family

The mention_priority_score is the anchor of our scoring framework. Rather than looking at a single metric like reach or raw engagement, MPS calculates a composite score from 0 to 100. It evaluates the historical authority of the author, the velocity of the engagement in the first two hours of publication, and the contextual density of the mention itself.

This score is highly calibrated. However, knowing that a mention scored an 82.4 while another scored an 82.7 is completely useless for a human operator. The human brain does not process third decimal places well when triaging a queue. The operator needs to know what action to take immediately, not the granular math behind the ranking.

This is where the bands come in.

The "Bands Not Points" Methodology

Bands translate raw mathematics into standard operating procedures. Instead of reading the exact score, you group scores into discrete bands and assign a rigid workflow to each band. Obsessing over individual points leads to analysis paralysis; managing by bands forces decisive action.

We routinely see teams succeed when they structure their scoring settings into four tiers:

• Critical Band (MPS 90 to 100): Existential threats, massive viral moments, or direct mentions from top-tier media publications. These require immediate, synchronous action.
• High Band (MPS 75 to 89): Significant conversations that require a response or escalation to the product team, but do not require waking anyone up at night.
• Moderate Band (MPS 50 to 74): General chatter, routine customer feedback, and low-level complaints. These are useful in aggregate but do not require individual operator attention on a daily basis.
• Low Band (MPS 0 to 49): Spam, passing references, and automated scraping noise. These are retained for volume analytics but completely hidden from human view.

By adopting this framework, you stop managing an infinite feed and start managing a discrete process.

Playbook: Setting Up the Workflow

To implement this methodology, you need to configure three specific product surfaces in your Sentia dashboard: your monitor settings, your routing integrations, and your inbox filters.

Step 1: Configure Scoring Settings

Navigate to your monitor settings. Under the scoring tab, you will find the scoring_settings JSON editor and the corresponding UI toggles. Here, you define the thresholds for your MPS bands.

Ensure that your entity definitions are tight. The MPS engine relies heavily on entity recognition to ensure the mention is actually about your brand and not a passing homonym. Once your entities are defined, set your band thresholds. We recommend starting with 90 for Critical and 75 for High, adjusting after your first week of data.

Step 2: Route Critical Alerts to a Telegram Channel

Critical mentions decay in value by the minute. If a prominent tech reviewer highlights a major bug about your product, your public relations and product teams need to know immediately.

Do not rely on email for the Critical band. Set up a webhook or use our native integration to pipe mentions with an MPS above 90 directly into a dedicated Telegram channel or Slack channel. The payload should include the author, the raw text, the URL, and a direct link to the Sentia Inbox for action. This creates a synchronous alert system for the things that actually matter.

Step 3: Triage the High Band in the Sentia Inbox

Your social operators should live primarily in the Sentia Inbox. Configure the default view of the Inbox to only show mentions that fall into the High band (MPS 75 to 89).

This is where the magic of the methodology becomes apparent. Instead of scrolling an infinite feed, the operator logs in, sees a tight, filtered queue of items for the morning shift, processes them, and achieves a zero-inbox state. They can reply directly from the interface, tag a colleague for review, or resolve the ticket and move on.

Step 4: Batch Process the Moderate Band

The Moderate band holds immense value for qualitative research, but it is a severe distraction for daily triage. Once a week, an analyst should review the Moderate band queue. They are not looking to reply to these users. They are looking for emerging trends · such as a sudden spike in complaints about a specific feature, or a new meme format taking hold in a niche community.

Step 5: Ignore the Low Band

Mentions below an MPS of 50 should practically never hit an operator screen. They exist purely for the charts on your macro dashboard. When the executive team wants to know the total share of voice, these mentions contribute to the pie chart. For the operator in the trenches, they do not exist.

A Realistic Scenario: The Product Launch

Let us apply this directly to the 14,310 scored mentions from our recent 30-day window. This occurred during an active mid-cycle product launch for a consumer brand.

Before Sentia, the brand's social manager spent hours a day manually reading forum threads and comments, trying to figure out what needed a response. They lived in a constant state of anxiety, worried they were missing something important in the massive flood of data.

After implementing the MPS bands methodology, the math completely changed their daily routine. Out of the 14,310 mentions:

• Critical (MPS 90+): A tiny fraction of the total volume. These triggered immediate alerts in the Telegram channel. The PR director handled them directly from their phone within minutes.
• High (MPS 75 to 89): A highly manageable daily subset. These populated the Sentia Inbox. The social manager cleared this queue swiftly each morning, responding to genuine customer questions and escalating bugs to support.
• Moderate (MPS 50 to 74): A broader segment of general market chatter. The marketing analyst reviewed a sample of these on Friday afternoons, noting a trend in requests for a new color variant.
• Low (MPS < 50): The vast majority of the sheer volume. Logged securely in the database, visualized on the leadership dashboard, and entirely ignored by the human operators.

Auditing Your Bands for Accuracy

Setting your thresholds is not a one-time task. As your brand grows or enters new markets, the baseline velocity of your mentions will naturally shift. A campaign that goes viral might temporarily skew your engagement metrics, pushing normal mentions down into lower bands.

To combat this, we routinely see operators perform a monthly band audit. The process is straightforward:

1. Open your Sentia Inbox and filter specifically for the Moderate band.
2. Sort the view by highest engagement rather than by MPS.
3. Look for mentions that clearly should have been classified as High or Critical.
4. If you spot anomalies, investigate the scoring_settings. You may need to adjust the weights of specific entities or increase the priority of particular authors in your allowlist.

Conversely, if your High band is filling up with irrelevant noise, your thresholds are too loose. Adjust the MPS threshold from 75 to 80 and observe the queue for a few days. The goal is a crisp, actionable inbox, and calibrating the bands is the only way to maintain that clarity over time.

The workflow we keep coming back to is one of radical exclusion. By trusting the calibrated impact score and ruthlessly enforcing the bands, the social team reclaims hours of deep work time. They stop drowning in metrics and start making decisions based on priority. The operator is no longer paralyzed by the sheer volume of the internet; they are empowered by a highly organized, highly actionable queue.