This case study examines how the City of Kawartha Lakes partnered with municiPal AI to improve resident access to municipal information. While the initiative’s initial goal was front-facing service improvement, implementation revealed a deeper problem: a “search tax” imposed on municipal staff across every department, every day. Drawing on documented ROI estimates and post-implementation observations, this case illustrates how information friction translates into measurable financial loss, and how targeted AI deployment can recover that cost.
Introduction
Governments at every level collect, maintain, and distribute vast quantities of information. Yet one of the most persistent and underappreciated operational problems in public administration is deceptively simple: employees and residents frequently cannot find the information they need, even when it exists.
This “information friction” is not a technology failure in the conventional sense. It is a structural condition produced by departmental silos, distributed content ownership, inconsistent documentation practices, and the sheer volume of digital content that accumulates in modern municipal operations. And it carries a cost.
This case examines how the City of Kawartha Lakes, a single-tier City in Ontario, Canada, in partnership with municiPal AI, a Canadian technology firm, set out to improve digital services for residents and inadvertently addressed a much larger problem.
Background and Context
The Municipal Information Environment
Public sector organizations manage information across dozens of departments, regulatory domains, and service channels. Unlike private firms that can enforce standardized knowledge management platforms, municipalities operate in highly federated environments where every municipality may maintain its own records systems, website content, policy documents, and procedural guides.
This challenge is not confined to the municipal level. The Government of Canada’s own digital strategy acknowledges that only 38% of federal government applications are considered “healthy,” and that the complexity of information flow across systems makes it difficult for people to find, navigate and use services. It’s a dynamic that affects internal staff as much as the public they serve (Treasury Board of Canada Secretariat, 2024).
About Kawartha Lakes
The City of Kawartha Lakes is a single-tier municipality serving approximately 75,000 residents across a large geographic area in central Ontario. Like many mid-sized Canadian municipalities, it provides a broad range of services — from property taxes and building permits to recreation programming and bylaw enforcement — through a workforce distributed across multiple service areas.
Residents had increasingly reported difficulty locating information online, encountering fragmented content or unclear pathways to relevant answers. Staff shared this frustration internally: routine tasks frequently required searching across multiple systems, confirming information across departments, and relying on informal or undocumented institutional knowledge.
About municiPal AI
municiPal AI is a Canadian technology company specializing in AI-powered search and information delivery for municipal websites. Its platform is designed to surface accurate, contextually relevant information to both residents and staff. Their solution replaces traditional keyword search with natural language understanding that can navigate the complex, document-heavy environments typical of government websites.
Defining the Search Tax
A search tax is the repeated expenditure of staff time required to locate information that already exists within organizational systems.
The concept of a “search tax” refers to the cumulative, compounding cost of time spent searching for information rather than applying it. While the phenomenon is recognized in organizational theory, it remains largely unmeasured in public sector environments, where its costs are distributed across hundreds of small daily interactions rather than appearing on any budget line or performance dashboard.
In Ontario, the conditions that make the search tax both invisible and expensive are well documented. Municipal information environments are characterized by departmental silos, where content is owned and updated independently across service areas, creating inconsistency between what exists and what is findable. Legacy content architecture accumulates years of bylaws, policies, and procedural updates that are rarely pruned or restructured. Staff frequently depend on colleagues rather than systems, creating key-person dependencies and interruption costs. And when residents cannot self-serve online, they generate call volumes that consume still more staff time which compounds the problem at both ends of the service relationship.
These structural conditions fall hardest on organizations that are already stretched. The Association of Municipalities of Ontario’s 2024 Municipal Workforce Development Project Roadmap found that many municipalities — particularly smaller ones — lack the staff time and financial resources to address known operational gaps, with one CAO capturing the condition plainly: “We know there are things we should be doing but we’re overwhelmed” (AMO, 2024, p. 5). StrategyCorp’s 2025 Ontario Municipal CAO Survey, based on interviews with 32 CAOs and City Managers across the province, reinforces this picture: local governments are being asked to “do more, for more people, with fewer resources and greater scrutiny than ever before” (StrategyCorp, 2025, p. 7). In this environment, the search tax carries consequences that extend well beyond administrative efficiency. Every minute a staff member spends locating information that already exists is a minute not spent on the service delivery, problem-solving, and community responsiveness that municipal leaders are under increasing pressure to demonstrate.
This challenge is not unique to the municipal level. The federal government’s 2025 budget explicitly identified internal inefficiency and duplication as drains on public service productivity, committing to technology and AI integration as the primary mechanism for recovery, not by adding capacity, but by reducing waste in existing processes (Government of Canada, 2025). The problem is not a shortage of staff or systems, but the friction embedded within them.
At Kawartha Lakes, that friction was not isolated to exceptional cases. It appeared during routine service delivery across departments, suggesting a systemic rather than situational condition. One that had never been formally identified or measured until an external intervention made it visible.
municiPal AI Implementation
Stated Objective
The primary objective of the Kawartha Lakes–municiPal AI partnership was to improve the resident-facing digital experience: simplifying how residents accessed municipal information, reducing inbound call volume, and increasing self-service task completion (permits, payments, program registration).
The expectation was that better organization and delivery of information would reduce inbound inquiries and improve self-service outcomes for residents. Internal efficiency was not an explicit design goal.
What Emerged
As the implementation progressed, a parallel challenge became visible: internal information retrieval was similarly fragmented. Staff engaging with improved information pathways found that the same structural problems affecting residents — inconsistent content, poor discoverability, reliance on informal knowledge — also governed their own daily workflows.
The initiative surfaced an internal inefficiency that had not been formally identified, measured, or budgeted for prior to implementation. The search tax had been invisible precisely because it was so ubiquitous.
Dual Outcome Model
The implementation ultimately produced two parallel streams of benefit:
- Resident experience improvements: Residents accessed clearer, more direct information through simplified digital pathways, reducing reliance on manual support channels.
- Internal operational efficiency: Staff experienced reduced need to manually search across systems to resolve routine inquiries. Information access became more direct and consistent.
Importantly, these improvements did not result from adding new processes or systems, but from reducing friction in existing information pathways.
Return on Investment Analysis
The following ROI estimates, drawn from municiPal AI’s analysis, apply conservative assumptions to illustrate the financial scale of the search tax problem. These figures represent recovered capacity, not new spending, which is a distinction with important implications for how public sector value is communicated to councils and budget committees.
All figures are modelled estimates based on publicly available workforce data and documented assumptions. They have been reviewed but not independently verified by the City of Kawartha Lakes and should be treated as illustrative of potential value rather than reported outcomes.
Note: All figures use fully loaded labour costs of approximately $40/hour unless otherwise stated. Estimates are conservative and based on municipal workforce benchmarks. Actual results will vary by municipality size and staff composition.
Customer Service Teams
Customer service staff represent the most direct intersection of the search tax and resident-facing service delivery. When staff cannot quickly locate accurate information, call handle times increase, escalations multiply, and resident satisfaction declines.
| Cost Category | Annual Value (10-person team) | Basis |
|---|---|---|
| Unlocked productivity (info lookup & email scripting) | $69,000 – $92,000 | ~1 hr/day redirected to higher-value work for a 10 person team @ $40 / hour, 230 days / year |
| Total | $69,000 – $92,000 |
General Staff — All Departments
The search tax is not confined to customer service. Every computer-based staff member who spends 5–10 minutes per day locating information is contributing to an organization-wide efficiency deficit. Across a typical municipal workforce, this compounding effect is substantial.
| Cost Category | Annual Value (per 100 staff) | Basis |
|---|---|---|
| Recovered information search time | $83,000 – $165,000 | 5–10 min/day at $40/hr fully loaded |
| Reduced errors and rework | $15,000 – $50,000 | 2 errors per department per month (assuming 5 departments handle bylaws, permits, policies, customer service) = 10 errors/month Each error takes on average 3 hours to identify, correct, communicate, and reprocess (one staff member) at $40/hour fully loaded |
| Reduced internal interruptions | $10,000 – $30,000 | 15 minutes per interruption × 2 staff members = 30 minutes of time lost per interruption At $40/hour fully loaded = $20 per interruption 500 interruptions per year (roughly 10 per week across the organization) = $10,000 |
| Faster onboarding (all departments) | $40,000 – $65,000 | Average fully loaded salary across all departments: $55,000/year $55,000 ÷ 52 weeks = $1,058/week 2 weeks of faster ramp-up = $2,115 per hire in recovered productive time |
| Total for 100-person team of computer-based roles | $148,000 – $310,000 |
For a municipality the size of Kawartha Lakes, with an estimated 630 computer-based staff (estimated to be 35% of total municipal staff), these figures suggest a recoverable search tax in the range of $930,000 – $1,950,000 annually, before accounting for councillor or communications team gains.
Elected Officials
Councillors operate under unique information demands: they must be prepared to vote on bylaws, respond to constituent inquiries, and participate in council discussions across a wide range of policy domains. The cost of inadequate information access at this level extends beyond productivity into reputational and governance risk.
| Cost Category | Annual Value (council of 10, 10 meetings / year) | Basis |
|---|---|---|
| Recovered councillor time | 440 hours / ~$25,000+ | 60 min/week x 10 councillors x 44 weeks. Assumes a conservative estimate of $50,000 annually for each. 5% of time recovered – $2,500 per councillor. |
| Deferred motions avoided | $20,000 – $100,000+ | 1 deferral per meeting × 10 meetings; $2,000–$10,000+ per deferral Cost per deferral assumes 11–26 combined staff hours to research, draft, review, and process a supplemental report. The lower range ($2,000) applies to routine administrative matters. The higher range ($10,000+) reflects deferrals requiring legal review, external consultation, or decisions with downstream procurement or project implications. |
| Council meeting readiness | $12,850 – $18,675 | Estimates for council meeting readiness aggregate three cost components: reduction in mid-cycle supplemental report requests (assumed 10 fewer per year at $1,000–$1,500 per report), reduction in post-meeting staff follow-up (assumed 30 fewer items per year at $45/hour), and reduction in in-meeting amendments requiring clerk and legal processing time. The upper range includes extended meeting costs where required staff attendance runs beyond scheduled session time. All figures assume a council of 10 with 10 annual meeting cycles. |
| Total for a 10 member council | $57,850 – $143,675 |
For a municipality like Kawartha Lakes, with council of 9, the combined value of recovered member time, avoided deferral costs, and improved meeting readiness is estimated at $52,000–$129,000+ annually, with the high end reflecting years where a complex deferral triggers legal review or delays a capital procurement decision.
Communications Teams
Municipal communications teams face a dual burden: they are responsible for the accuracy and accessibility of public-facing content, and they are frequent victims of the search tax when coordinating across departments. AI-assisted information management can reduce both content maintenance overhead and the cost of misdirected campaign spend.
| Cost Category | Annual Value (team of 3–5) | Basis |
|---|---|---|
| Content management savings | $11,000 – $28,000 | Assumes a comms team of 3–5 staff recovering 2–3 hours per week previously spent auditing stale content, fixing broken links, chasing departmental contributors, and manually tracking update schedules @$40 / hour. |
| Smarter campaign spend | $10,000 – $40,000 | Assuming a municipality with a $50,000 – $200,000 annual communications budget avoids 20% in misdirected spend = $10,000 -$40,000 in annual savings. |
| Fragmented Internal Systems | $13,800 – $46,000+ | A team of 3 spending 30 min/day searching internal systems = $13,800. A team of 5 spending 60 min/day = $46,000. |
| Tool consolidation | $8,000 – $20,000 | One platform for updates, alerts, communications |
| Total for a 3-5 person comms team | $42,800 – $134,000 |
For Kawartha Lakes’ Communications team of 4, estimated annual value ranges from $51,000–$119,000, driven primarily by time recovered from content maintenance and internal systems searching, smarter campaign spend informed by real resident search data, and consolidation of overlapping tools.
Modeled Savings
Taken together, these estimates suggest that the City of Kawartha Lakes — with approximately 630 computer-based staff, a customer service team of 10, a council of 9, and a communications team of 4 — could conservatively be losing between $1.1 million and $2.3 million annually to the search tax across all departments and levels of the organization.
That figure does not represent new spending. It represents existing staff time, organizational capacity, and budget that is currently absorbed by the friction of finding information that already exists. Recovering even a fraction of it, through faster information access, fewer deferred motions, reduced content maintenance overhead, and smarter campaign spend, represents a meaningful return on a modest technology investment.
Organizational and Strategic Implications
The Visibility Problem
One of the most instructive aspects of this case is that the search tax was not unknown but it was unmeasured. Staff and managers were aware that finding information was time-consuming, but this friction had never been quantified, reported, or included in operational planning.
This invisibility is common in public sector settings, where efficiency losses that are distributed across many small daily interactions rarely appear on any dashboard or budget line. Unlike a failed system or a budget overrun, the search tax happens quietly, compounding drain on organizational capacity.
The Dual Benefit Model
This case demonstrates that external service delivery improvements and internal operational efficiency are not separate agendas. When information becomes more accessible to residents, it often becomes more accessible to staff simultaneously, because the underlying problem (poor information architecture) is the same.
This has strategic implications for how municipalities frame technology investment proposals. A case built solely on resident-facing outcomes may understate the true ROI by half or more. CFOs and CAOs evaluating AI investments should be asked to account for internal efficiency gains, not only service delivery metrics.
A persistent challenge in municipal finance is how those gains are classified. Recovered staff time and avoided costs are frequently categorized as “soft savings” which are acknowledged informally but excluded from formal ROI calculations because they do not reduce a specific budget line. Yet, if municipalities only ever count what shows up as a line reduction, they will systematically undervalue (and underinvest in) tools that make their organizations measurably more effective.
Conclusion
The Kawartha Lakes–municiPal AI case reveals a quiet and pervasive cost that governments rarely measure or discuss: the ongoing expenditure of staff time in search of information that already exists.
The search tax is real, it is large, and it is recoverable. Conservative estimates suggest that a mid-sized municipality could be losing hundreds of thousands of dollars annually to information friction. That is before accounting for governance risk, reputational cost, or resident experience degradation. This is not a technology problem. It is a management problem that technology can help solve.
This case offers a model for how to identify hidden operational costs, build evidence-based investment cases, and evaluate the organizational conditions that determine whether technology interventions succeed or fail. It also invites reflection on a question that the case study itself raises: why does it have to be so hard to find information that the government already has? Increasingly, it doesn’t.
References
Treasury Board of Canada Secretariat. (2024). Canada’s digital ambition 2024–25. Government of Canada. https://www.canada.ca/en/government/system/digital-government/canada-digital-ambition/canada-digital-ambition-2024-25.html
municiPal AI. (2025). A smarter way to serve: Kawartha Lakes’ AI Driven Website Transformation https://municipalai.ai/a-smarter-way-to-serve/
municiPal AI. (2026). Is Your Municipality Paying The Search Tax? https://municipalai.ai/is-your-municipality-paying-the-search-tax/
Association of Municipalities of Ontario. (2024). Careers that build communities: AMO’s municipal workforce development project roadmap. Association of Municipalities of Ontario. https://www.amo.on.ca/sites/default/files/assets/AMOWorkforceDevelopmentProjectRoadmap2024.pdf
StrategyCorp. (2025). Ontario municipal chief administrative officer survey: A candid look at the issues on the minds of Ontario’s CAOs. https://strategycorp.com/wp-content/uploads/2025/08/StrategyCorp-CAO-Report-2025-08-1512.pdf?utm_source=chatgpt.com
Government of Canada. (2025). Budget 2025: Chapter 5 — Creating a more efficient and effective government. https://budget.canada.ca/2025/report-rapport/chap5-en.html
Appendix A: Summary ROI at a Glance
Appendix A: Estimated Annual Search Tax by Team
Municipalities can use the tables below to build a locally relevant estimate. Identify your approximate staff count or team size in each table and add the figures together for a combined organizational estimate.
Table 1 — General Staff (computer-based roles)
5–10 min/day recovered at $40/hr, plus reduced errors, interruptions, and faster onboarding
| Computer-Based Staff | Conservative | High-End |
| 50 | $74,000 | $155,000 |
| 100 | $148,000 | $310,000 |
| 150 | $222,000 | $465,000 |
| 200 | $296,000 | $620,000 |
| 300 | $444,000 | $930,000 |
| 400 | $592,000 | $1,240,000 |
| 500 | $740,000 | $1,550,000 |
| 600 | $888,000 | $1,860,000 |
| 750 | $1,110,000 | $2,325,000 |
| 1,000 | $1,480,000 | $3,100,000 |
Table 2 — Customer Service Team
1 hr/day at $40/hr, redirected from information lookup to higher-value resident interactions x 230 days / year
| Team Size | Conservative | High-End |
| 5 | $34,500 | $46,000 |
| 8 | $55,200 | $73,600 |
| 10 | $69,000 | $92,000 |
| 12 | $82,800 | $110,400 |
| 15 | $103,500 | $138,000 |
| 20 | $138,000 | $184,000 |
Table 3 — Council
60 min/week recovered (46 active weeks, $50K/year), 10 meetings/year, 1 deferral per meeting at $2,000–$10,000+, meeting readiness savings
| Councillors | Conservative | High-End |
| 5 | $45,350 | $131,175 |
| 6 | $47,850 | $133,675 |
| 7 | $50,350 | $136,175 |
| 8 | $52,850 | $138,675 |
| 9 | $55,350 | $141,175 |
| 10 | $57,850 | $143,675 |
| 12 | $62,850 | $148,675 |
| 15 | $70,350 | $156,175 |
| 20 | $82,850 | $168,675 |
Table 4 — Communications Team
Content maintenance and internal systems
2–3 hrs/week content maintenance + 30–60 min/day internal systems searching at $40/hr
| Team Size | Conservative | High-End |
| 3 | $24,840 | $44,160 |
| 4 | $33,120 | $58,880 |
| 5 | $41,400 | $73,600 |
| 6 | $49,680 | $88,320 |
| 8 | $66,240 | $117,760 |
| 10 | $82,800 | $147,200 |
Consolidated tools and campaign spend
Assuming the communications budget avoids 20% in misdirected spend
| Conservative | High-End | |
| Smarter campaign spend | $10,000 | $40,000 |
| Tool consolidation | $8,000 | $20,000 |
| Total | $18,000 | $60,000 |
Note: Add figures from each applicable table to produce a combined organizational estimate. All figures assume $40/hour fully loaded unless otherwise stated. Governance and reputational risk values are excluded from all totals.