For brand agencies & consultancies

Make customer correspondence
your next consulting offer

Your clients spend millions building brands - then undermine them one email, letter, SMS and chatbot response at a time. CommScore.AI is a communications optimisation platform that measures each message against brand and communications principles - and shows how to close the gap.

5
communication quality pillars: Clarity · Certainty · Structure, Style & Grammar · Brand Alignment · Customer Experience
30
quality checks on every communication scored
20+ years
of communications experience in the making
The CommScore.AI platform in three steps: input your brand or service values, upload messages for review, then get recommendations for optimising experience and outcomes - shown as a scored report with an overall CommScore and per-pillar detail.
Add a brand, upload the messages, then read the score and the prioritised recommendations.

CommScore.AI is evidence-based consulting

Every message is scored from one to five, across five quality pillars.

Instead of saying: “We think this communication could be better.”

You can say: “We’ve measured it. This is what’s holding it back. Here’s a better version. And the improved score to prove it.”

Report overview
Renewal notice · Email Needs attention

Please be advised that your enquiry is being actioned.

2.4/ 5
  • Clarity 2.8
  • Certainty 2.0
  • Structure, Style & Grammar 3.4
  • Brand alignment 1.6
  • Customer experience 2.6

Brand alignment is the weakest pillar - and a weak pillar caps the overall score. It is the fastest thing to fix.

Same notice, rewritten Strong

We’re on it. You’ll hear from us by Thursday.

4.6/ 5
Illustrative CommScore.AI report. The “before” sentence is the kind of phrasing a brand’s tone guidance exists to prevent.

Every communication becomes progressively better

Clearer, more consistent, more effective - and you can show the client exactly how much, on every pass.

The CommScore.AI continuous optimisation cycle: Review, Recommend, Rewrite, Rescore, repeating.
  • 1 Review CommScore.AI reviews the communication across five key pillars and 30 criteria.
  • 2 Recommend CommScore.AI identifies gaps, highlights opportunities and provides clear, prioritised recommendations.
  • 3 Rewrite Use the recommendations to rewrite for greater clarity, effectiveness and brand fit.
  • 4 Rescore Rescore the rewritten message to measure improvement and prove impact.

Consulting revenue hiding in plain sight

Every organisation has hundreds - sometimes thousands - of templated customer communications. Most have never been reviewed. That is a body of work waiting to be scoped.

Health check

Communications Health Check

Review and score 10–50 communications.

Typical engagement $10k – $30k
  • A baseline CommScore for each communication
  • Ranked recommendations, pillar by pillar
  • An exportable report for the client’s board
Transformation

Communications Transformation

Review, rewrite and optimise 200+ communications.

Typical engagement $100k – $1M+
  • The whole portfolio scored and prioritised
  • Rewrites, scored before and after
  • The weakest communications surfaced first
Continuous

Continuous Optimisation

Ongoing monitoring, benchmarking and improvement.

Typical engagement From $5k per month
  • New communications scored as they ship
  • Re-score on a cadence you set
  • Evidence for the next renewal conversation

Indicative engagement values in Australian dollars, drawn from Customer Science’s own delivery experience. What you charge is yours to set.

Why not just ask ChatGPT or Claude?

You absolutely can - and for one-off writing tasks, they are excellent. The difference is that CommScore.AI is not a writing assistant. It is a communications assessment and optimisation platform.

Where a general-purpose AI gives you an opinion shaped by your prompt, CommScore.AI applies the same scoring framework to every communication - Clarity · Certainty · Structure, Style & Grammar · Brand Alignment · Customer Experience - drawn from the same thirty checks every time. That consistency is the point, and it is what makes the output evidence rather than advice.

  • Score communications consistently across teams, business units and time.
  • Benchmark quality before and after improvements.
  • Score at scale - hundreds or thousands of communications, judged the same way each time.
  • Apply your own brand and service principles, not generic writing advice.
  • Identify specific opportunities for improvement, rather than simply generating a rewrite.
  • Demonstrate measurable improvement to stakeholders.
  • Optimise your most essential communications channel - the one your customers actually read.

More questions

Will this replace our writers?

No. It makes good writers more effective, and helps non-writers produce better communications.

What if our communications are already pretty good?

Then you will have objective evidence that they are good - and you will be able to identify the remaining opportunities, which are the hardest ones to see.

Can AI really judge writing?

AI does not replace human judgement. CommScore.AI applies a structured assessment methodology, developed from decades of customer communications expertise, to identify issues consistently and at scale.

You stay in the loop by design: before anything is scored, you confirm the text we read, the communication type, who it is for, and whether the reader is expected to act.

Why does scoring matter?

Without scoring, improvement is largely subjective.

With scoring, you can measure progress, prioritise effort, benchmark quality and demonstrate return on investment.

Your clients’ communications stay inside the boundary

The communications you score may contain personal information - a customer’s email, a claim, a bill. CommScore.AI is built so that content never leaves the Azure boundary, is never kept longer than it must be, and is never used to train a model.

The original file is never stored

Uploads are extracted to text and the binary is discarded. Only the text you confirm is kept.

Your data does not train models

All scoring runs on Azure OpenAI under Microsoft’s data-processing terms. Prompts and completions are not used to improve any model, and are not reviewed by humans.

Prompt content is never logged

AI request logging is limited to non-identifying context - model and basic parameters. Instructions and messages are not written to logs.

Nothing is scored without you

You confirm the extracted text before analysis. There is no path that scores a raw file you have not seen.

Read the security overview → Customer Science Group holds ISO 27001:2022 certification for information security management.

Twenty years of customer communications, across the brands Australians deal with every day

  • Alinta Energy
  • Allianz
  • ANZ
  • Australian Taxation Office
  • Australian Unity
  • Bank of Melbourne
  • BT Financial
  • HCF
  • Macquarie
  • Medibank
  • Momentum Energy
  • NAB
  • nib
  • QBE
  • Telstra
  • Velocity Frequent Flyer

The communications expertise behind CommScore.AI was built on two decades of transformation work with Australia’s largest service organisations - in banking, insurance, health, energy, government and travel. That experience is what the five pillars and 30 checks encode.

Sell evidence, not opinions.

Bring one of your client’s communications to the demo. We’ll score it with you.