Founder's Take

Score finds it. Forge builds it.

Score finds it. Forge builds it.

Your findings deserve a better ending than the backlog.

Your findings deserve a better ending than the backlog.

- Dominic Francis

Co-founder, fwdDeploy

Dominic Francis, Founder's Take on the Freshservice health check: Score finds it. Forge builds it.

Table of Contents

I've seen a lot of service desk assessments. Most end the same way. A PDF full of gaps and scores lands in someone's inbox. Then nothing happens. 

So we built Score inside Forge, our AI configuration accelerator. Score runs a Freshservice health check on your live instance. Actionable configuration findings can flow into Forge for review and implementation. Then you scan again and watch the number move. 

That loop is the whole point. Here's how it works, and why I think it changes the job of a service desk admin. 

Why every service desk assessment hits the same dead end 

Most assessments end with a report, not a fix. The PDF of gaps and scores gets filed. The findings go to the backlog, and the backlog goes to next quarter. Six months later, nothing has changed. 

Too many service desk assessments end with recommendations that never get implemented.

— Dominic Francis


The assessment did find the problems. Fixing them takes configuration work, and the person who owns Freshservice is already busy running the service desk. So the report waits. 

Across years of ITSM implementations and optimization engagements, our team has repeatedly seen the same pattern: identifying opportunities is often easier than finding the time and resources to implement them. The software was never the blocker. Getting it fully deployed was. 

What a Freshservice health check should actually measure 

A useful health check should go beyond identifying what's configured in your Freshservice environment. It should answer three important questions: 


01

What's
configured?

02

What's actually being used?

03

How well is the desk performing?


Score, our Freshservice health assessment capability within Forge, brings these three perspectives together. It evaluates configuration, examines actual ticket activity, compares service performance against industry benchmarks, and identifies opportunities to get more value from the platform. 

More importantly, it helps distinguish between a service desk that's delivering good results and one that's operating efficiently. 

Strong results don't always mean an optimized service desk 

Consider a sample Freshservice environment with 26 agents, running on the Enterprise plan. Score analyzed its configuration and a sample of 120 tickets. 

The environment received an overall Health Score of 77%, combining a 74% configuration setup score with an 80% service performance score. 

At first glance, here's how the service desk was performing

97.4%

Resolved on time

Benchmark 95%

97.4%

Resolved on first contact

Benchmark 72%

4.8h

Average resolution time

Benchmark 25.6h

5 of 6

Indicators above industry average

Service performance

Source: https://www.freshworks.com/freshservice/benchmark-report-2025/

But the assessment uncovered a different story when it examined how those results were being achieved. 

The service desk was performing well, but much of the work was still being done manually. 

Only 27 of the 120 sampled tickets (22.5%) showed evidence of automation. Of the 39 workflow automation rules configured, just 19 were enabled. The remaining 20 were switched off or had never been activated. 

Despite having 26 agents, the environment had no ticket templates or checklist templates configured. 

Finding opportunities hidden in existing configuration 

Score also examines how individual Freshservice capabilities are configured and used. 

In this example, the service catalog was performing well. It contained 27 catalog items, and 60.1% of requests came through the catalog, exceeding the assessment's 40% target. 

However, none of those 27 items had portal descriptions. Similarly, the environment had nine agent groups, with 95.7% of tickets assigned to a team. Yet the average time to assign a ticket was still 9.5 hours, suggesting an opportunity to improve routing and assignment efficiency. 

Despite strong on-time results, the environment relied on a single SLA policy covering every priority. Score recommended reviewing whether differentiated SLA targets would better reflect urgent, VIP, and business-critical work. 

A capability can be configured and functioning without necessarily being optimized for the organization's needs.

— Dominic Francis

Score also flags where evidence is missing, so saved replies, workflow execution counts and knowledge management were marked for further review instead of being scored as failures. 

Configured vs optimized table for Freshservice catalog, agent groups and SLA policies, with what works and the gap


Turning findings into a practical 90-day improvement plan 

Identifying opportunities is only the beginning. Here's what the 90-day plan looks like:

Days 0–30

Establish the foundation

Review the 20 inactive or never-activated automation rules, assign ownership and add descriptions to all 27 catalog items.

Days 31–60

Routing & SLAs

add routing and auto-assignment, differentiated SLA policies for priority and VIP work, and required impact and location fields.

Days 61–90

Optimize productivity and prepare for AI

That means ticket templates and checklists, Active Directory, Azure and Slack integrations, and a Freddy AI Copilot pilot with a selected team. It also adds CSAT measurement and a follow-up health assessment. Longer term, the roadmap points to Freddy AI Agent, plus problem and asset management.

From findings to built configuration 

This is where Score becomes an agent. Every approved roadmap item flows back into Forge's AI configuration pipeline. 

Forge already turns discovery notes and call transcripts into built Freshservice environments. Score adds live configuration evidence to Forge's implementation process. 

From each finding, Forge produces a ready-to-review configuration template. Deterministic validation checks the proposal first. Then your team and your Forward Deployed Engineer (FDE) review it. You review. You approve. Forge configures it. 

Scan

→

Findings

→

Roadmap

→

Approved fixes

→

Built

Governance is built into every step. The AI recommends; controlled workflows execute. Here's why we chose governed AI over autonomous AI. 

Forge also reviews each design against ITSM practice before it builds. It builds twelve categories of Freshservice configuration in minutes. When a fix needs more than a template, our FDEs own the buildout and the support after go-live. 

We've completed more than 25 Freshservice assessment and configuration engagements using Forge. Configuration work that used to take many days now takes under an hour.

Customer result

"For Crystal Clean Freshservice project, we used Forge to clone configurations from one workspace to another, including the required setup and configuration.

What typically takes an estimated 1-2 days of manual effort, including configuration, testing and validation was completed in approximately 5 hours using Forge.

This significantly improved delivery speed and reduced the overall implementation effort."

Why service desk health is ongoing upkeep 

A service desk keeps changing. New agents join, and rules that worked last quarter go stale. One clean assessment can't hold that still. 

Service desk health is upkeep, not a one-time project.

That's why the roadmap establishes measurable targets for a follow-up assessment. For this sample environment, the proposed day-90 outcomes include increasing automated ticket handling from 22.5% to 70%, moving from one catch-all SLA policy to three or more differentiated policies, increasing the proportion of enabled automation rules from 49% toward 90%, and ensuring all catalog items have meaningful descriptions. 

Freshservice health check day-90 targets table: automation 22.5% to 70%, SLA policies 1 to 3+, enabled rules 49% to 90%

These targets provide a basis for evaluating progress. They are not guaranteed outcomes, and a higher score alone does not prove that every change improved service quality. 

So you run the Freshservice health check again and watch the trend line. The real value comes from comparing configuration changes with operational results over time. 

Admins can spend less time on repetitive configuration and more time reviewing decisions and validating outcomes. The weeks of implementation move to Forge and your FDE. Then scan again. 

Cycle diagram: Scan, Findings, Roadmap, Approved Fixes, Built, then Rescan back to the start

That's what a Freshservice health check should deliver: not simply a score, but evidence-backed recommendations, a practical improvement plan, and a way to measure whether the changes are making a difference. 

Freshservice first, with a platform-agnostic core 

Score starts with Freshservice. It reads your plan, your modules and your Freddy AI entitlements. 

Underneath, the architecture is platform-agnostic. Score already exists for Rocketlane, where it scores 16 implementation dimensions. I covered that in my Forge for Rocketlane post. 

See what a scan finds in your instance

Too many good assessments die in shared drives. That's why Score exists. Your findings deserve a better ending than the backlog.

Run a scan on your own Freshservice instance. Put what's configured next to what's actually used, and notice where the two drift apart. A Score scan is one simple way to do that.

I'd like to hear what your scan turns up. Send me the finding that surprised you most, and let's keep the conversation going.

Authored By

Dominic Francis

Co-founder, fwdDeploy

Dominic co-founded fwdDeploy with a simple objective: help organizations implement, integrate, migrate, and optimize their enterprise technology platforms more effectively.

See what a scan finds

Run Score on your Freshservice instance and get a 30/60/90-day roadmap you can approve and build.