PROVENRESULTS
Numbers tell the story better than we can. Here are selected engagements — anonymized where required — that illustrate how we work and what we deliver.
Campaign Controls & Billing Recovery
Top-3 Canadian Telecom Provider
The Problem
Over 200 active promotional campaigns had been released without adequate controls, causing billing inaccuracies affecting 100,000+ customer accounts. Customers were overcharged, undercharged, or left on expired promotions — driving complaint calls and disconnections.
Our Approach
We analyzed every active campaign against the billing system, identified 220 broken configurations, and quantified the revenue exposure per campaign. We then designed a term-offer conversion framework with built-in eligibility controls, CSR-level testing requirements, and automated validation before launch.
The Result
$56M in total identified impact — $2.6M from call volume reduction and $54M from preventing customer disconnections. Corrective actions applied across affected accounts up to 3 months in arrears.
Tech Note
Data extraction from billing platforms, statistical analysis of campaign attachment rates, root-cause mapping across product/promotion combinations.
Call Center AHT Reduction
Major Canadian Cable & Wireless Provider
The Problem
Average Handle Time across cable, common, and wireless call centers was significantly above targets. Leadership suspected systemic issues but lacked visibility into root causes across 3 lines of business.
Our Approach
We performed agent-level behavioral analysis — tracking patterns of blind transfers, excessive after-call work, auxiliary time abuse, and muted hold times. We segmented opportunities by call center type and built a prioritized action plan targeting behavior-driven AHT inflation.
The Result
Identified 30–60 seconds of reducible AHT per call, representing $14M–$40M in annual OPEX savings weighted heavily toward wireless. Delivered alongside a billing error root-cause analysis that uncovered the source of 42% of billing complaint calls.
Tech Note
Agent-level CMS data mining, statistical decomposition of AHT components, behavioral pattern clustering.
Address Inventory Modernization
National Telecom Network Operator
The Problem
Address inventory records used in FTTH planning varied from actual field results by 10–30%. The incumbent vendor quoted $1.96M and 4–6 months for a conventional migration. Address data inconsistencies between systems were actively delaying network builds.
Our Approach
We proposed an alternative: agile development with a collaborative data modeling approach using the client's existing datasets. We aligned address sources, rebuilt connectivity models, and replaced the vendor's waterfall plan with an iterative 2.5–3.5 month program.
The Result
Delivered for $390K — an 80% cost reduction against the vendor quote — in under 3.5 months. The solution included reusable ETL pipelines and alignment of address records across FMS, LPDS, and the GIS platform.
Tech Note
SpatialNET data modeling, address source reconciliation, FME-based ETL pipelines, custom Python validation tools.
IT Asset Inventory Optimization
Major Canadian Bank
The Problem
IT hardware inventory was growing uncontrollably — 80% of ordered assets sat in storage for over 2 months before consumption, and increasing SKU variability meant more equipment was ordered than consumed. Storage costs and obsolescence risk were escalating.
Our Approach
We mapped the full asset lifecycle — from order date through physical arrival, consumption, and end-of-life — across all product types. We identified three compounding problems: long lead times (2.5 months), high arrival-to-consumption gaps (54+ days), and growing inventory from demand variability.
The Result
Recommended and designed a 50–70% reduction in inventory stores through improved demand forecasting, SKU rationalization, and supplier lead-time management. The analysis revealed that 15% of average asset lifespan was consumed in the storage process alone.
Tech Note
AMS data mining across 3 years of order/arrival/consumption records, demand variability analysis, lifecycle modeling.
Data Mining for Churn & OPEX Reduction
Leading Fleet Telematics Platform
The Problem
The client needed to reduce operational costs by 25% in the fulfillment-to-support process while simultaneously reducing customer churn. Existing reporting was retrospective and disconnected from the user interfaces where decisions were made.
Our Approach
We reverse-engineered data models across Salesforce, F&O, and the web platform, then co-validated process views through historical data. This approach bypassed months of requirements gathering and went straight to insight. We identified gaps between what systems captured and what operators actually needed to see in real time.
The Result
Quickly enabled the capture of $1.6M–$3.2M in annual churn. Delivered a retrospective framework that highlighted the gap between historical reporting and the real-time system views needed for proactive decision-making.
Tech Note
Cross-platform data mining (Salesforce, F&O, 8x8), Python data pipelines, behavioral churn modeling, process gap analysis.
Automated GIS Migration Quality Assurance
North American Cable Operator
The Problem
A multi-year GIS migration from legacy Microstation and GE Smallworld platforms to a modern enterprise application required quality assurance across hundreds of thousands of spatial records — splice cans, fiber routes, coaxial plant, and connectivity data.
Our Approach
We extracted and simplified source data from Microstation, built automated QA validation pipelines, and deployed migration quality assurance tools that could verify data integrity at scale. Separately, we automated splice can report generation to eliminate the need for a $20K+ legacy license upgrade.
The Result
Delivered automated QA tooling covering multiple database instances across two countries. The splice report automation was delivered at $0 (goodwill) and replaced a vendor proposal that would have cost $20K+ for a single named license.
Tech Note
FME data extraction, GE Smallworld to PostgreSQL/PostGIS pipelines, Python-based automated QA scripts, Google Cloud storage integration.
Ready to talk?
Every engagement starts with a conversation. Book a free 25-minute consultation to discuss your challenges and explore how Tacit Insight can help.
Book a Consult