InsuredMineAI Pulse Report 2026

AI writes the email. People still key the quote. Built for agency owners, principals, strategists, and technologists: see where you stand on AI, where to start, and how to stay ahead.

Download the report Talk to us

Scroll to read

Why this survey

Adoption is easy to count. What agencies need from AI is harder to see.

What agencies need from AI, and what stands in their way, is the question behind this survey. InsuredMine sits where an agency’s people, its agency management system (AMS) and its clients meet, which gives us a close view of where work flows and where it stalls. We asked our customers directly, across every agency size, about the gaps, problems, demands, cautions and roadblocks they face. What they told us sets our agenda: to resolve those problems and to help every agency we serve keep its competitive edge.

Thank you!

To every InsuredMine customer agency that took part: this report exists because of the time and candor you gave us. It would not have been possible without you.

With gratitude, the InsuredMine team

Executive summary

The 5 key findings

82 InsuredMine customer agencies answered 13 questions between mid-July and mid-September 2026. At least 4 in 5 write both personal and commercial lines. They chose to respond, so the results describe this sample, not the market, and may skew more engaged with AI.

Adoption is broad and shallow. AI is in use at 84% of agencies, and 52% stop at email and documents or have not started.

Your first movePick one repeated workflow.

The hours are real and uncounted. 22% lose more than 10 hours a week, and 20% cannot say.

Your first moveCount them for 2 weeks.

Readiness runs ahead of data flow. Large agencies rate themselves the most ready and their data flow among the lowest, and agencies that name 3 or more broken integration seams are about twice as likely to lose 5 or more hours a week.

Your first moveTrace one record from CRM to AMS.

Owners want proof. 34% ask for workflow clarity, and 4% ask for an ROI model.

Your first moveHave any vendor show a real workflow.

The first move depends on size. Small, midsize and large agencies are stuck in different places.

Your first moveUse the 30-day plan for your size.

Nuances of note

Four further points shape how to read these findings. Privacy and compliance concern holds near 40% at every agency size; what rises with size is the sense of being past the barriers, from 8% of small agencies to 45% of large. Intent runs ahead of use: 15% use predictive analytics or renewal risk scoring today, and 35% plan to adopt automated renewal risk scoring. The strong preference for native integration (54% call it a must) should be read with care, because the question set it against middleware. Finally, the midsize and large groups are small, so those patterns are directional. The common thread is a request to see one workflow improve on the agency’s own data.

Want to see where your agency sits?

Bring one workflow. We can look together at where the work happens, what your systems carry across, and what is worth measuring first.

Talk to our team

Chapter 1 of 4

Adoption is broad but shallow

Agencies want to be AI-ready, and most say they are not there yet

The average agency rates its readiness to adopt AI at 2.85 out of 5. Only 34% rate it 4 or 5, and 39% rate it 1 or 2. Readiness climbs with size, from 2.60 at small agencies to 3.91 at large ones. Low readiness is not the same as standing still: of the 32 agencies that rate themselves 1 or 2, 21 (66%) still name an AI capability they expect to adopt within 12 months. The distance between where agencies are and where they want to be is wide, and most intend to close it.

Exhibit 1

Readiness rises with agency size

Q1How would you rate your agency’s current readiness to adopt AI-powered tools? (1 = just getting started, 5 = AI is part of our operations)

34%of agencies rate themselves 4 or 5 (small 28%, midsize* 40%, large* 64%)

39%rate themselves 1 or 2 (small 47%, midsize* 27%, large* 0%)

66%of the 32 agencies rating themselves 1 or 2 still name a capability to adopt within 12 months

All 28 agencies that rate their readiness 4 or 5 also name an AI capability they plan to adopt in the next 12 months.

* Small samples; read directionally. Sample sizes: small n = 53, midsize n = 15, large n = 11. Average uses those answering; 1 respondent did not answer. Source: InsuredMine AI Survey, 2026.

Use is broad but shallow

AI is in use at 84% of agencies, concentrated in the least demanding work: 72% draft email and 56% summarize documents. Beyond those two it falls, to 32% for data-entry or workflow automation, 21% for automated client follow-up and 15% for predictive scoring. Overall, 52% have deployed nothing or stop at email and documents. Size sets the pace: 66% of small agencies sit at that level, against 27% of midsize and 18% of large.

Exhibit 2

AI use is concentrated in email and documents

Q2Which AI applications is your agency currently using? (multiple selections)

Email drafting runs from 58% of small agencies to 100% of large. Workflow automation runs from 19% to 64%.

Our reading. Drafting and summarizing sit beside the work and ask little of the rest of the business. Automation, follow-up and scoring need accurate data in the right place. That is our interpretation, not a survey finding.

* Small samples; read directionally. 3 respondents did not report size. Source: InsuredMine AI Survey, 2026.

Trust, not budget, holds agencies back

Privacy, security and compliance concerns (43%) and not knowing which use cases are worth prioritizing (39%) lead the barriers. Budget (13%) and leadership priority (10%) sit at the bottom. Privacy concern holds near 40% at every size, while the sense of being past the barriers rises from 8% of small agencies to 45% of large. One barrier points past AI itself: 27% say their systems do not integrate well enough to make AI useful.

Exhibit 3

Privacy is the common concern; prioritization holds back small and midsize agencies

Q3What is the biggest barrier to AI adoption at your agency? (worded “biggest”; multiple selections collected)

* Small samples; read directionally. 1.7 selections per respondent. Source: InsuredMine AI Survey, 2026.

Chapter 2 of 4

Readiness runs ahead of data flow

Large agencies rate themselves the most ready, and their data flow among the lowest

Readiness and data flow part ways at the top. 64% of large agencies rate their readiness 4 or 5, while 18% rate their data flow that high. Across all agencies, data flow averages 2.86 out of 5, and only 4 (5%) call it fully automated. The split shows up agency by agency, too: of the 28 agencies that rate their readiness 4 or 5, 7 (25%) rate their data flow 4 or 5.

Exhibit 4

Readiness rises with size. Data flow does not follow.

Q4How would you rate the quality of your current CRM ↔ AMS data flow? (1 = mostly manual, 5 = fully automated). Readiness is Q1, Exhibit 1.

Average readinessAverage CRM–AMS data flow

Small

28%rate readiness 4 or 5

25%rate data flow 4 or 5

Midsize*

40%rate readiness 4 or 5

40%rate data flow 4 or 5

Large*

64%rate readiness 4 or 5

18%rate data flow 4 or 5

Our reading. At the largest agencies, confidence has outrun the connection underneath it. The survey does not show that weak data flow causes shallow adoption; it shows that the two part ways.

* Small samples; read directionally. Size bands: small 25 or fewer employees, midsize 26 to 100, large 101 or more. Averages use those answering; 1 respondent did not answer Q1 or Q4. Source: InsuredMine AI Survey, 2026.

The friction is specific: the same few seams break again and again

Only 2% of agencies say their CRM and AMS do not connect at all, and 9% run no separate CRM. The friction is in what fails to carry across. Activity logging (38%) and contact and account sync (37%) are named most often, one respondent apart. The pattern shifts with size: reporting reconciliation runs from 13% of small agencies to 60% of midsize, and contact sync peaks at 64% of large agencies. 70 of 82 agencies answered this question. The next chapter shows why the number of broken seams matters.

Exhibit 5

Activity logging and contact sync are the most-named gaps, and the pattern shifts with size

Q5Which integration gaps cause the most friction for your team? (multiple selections)

* Small samples; read directionally. 70 of 82 answered; 12 did not. Source: InsuredMine AI Survey, 2026.

See where your records break

Calls are logged automatically and records sync both ways with your AMS. We’ll trace a record with you.

Talk to our team

Chapter 3 of 4

The hours are real and uncounted

The hours are real, and 1 in 5 agencies cannot count them

39% of agencies lose 5 hours or more a week to manual data entry that could be automated, and 22% lose more than 10. Another 20% have never measured it. By size, 19% of small agencies, 40% of midsize and 18% of large lose more than 10 hours a week, and 45% of large agencies have not counted at all. An agency that cannot say what the work costs cannot choose where to start.

Exhibit 6

1 in 5 agencies loses more than 10 hours a week to manual entry

Q6How many hours per week does your team spend on manual data entry that could be automated?

Source: InsuredMine AI Survey, 2026.

What the hours are worth Illustrative · modelled

Among the agencies that gave a figure, the average is about 5.9 hours a week, using band midpoints and counting “more than 10” as 10. These figures are modelled from time bands. They are not reported dollars or promised savings.

~300 hoursa year at the average of 5.9 hours a week

$7,500 to $15,000a year at an assumed loaded labor cost of $25 to $50 an hour

520+ hoursa year, or $13,000 to $26,000, at agencies above 10 hours a week

Quoting and renewals take the week

Quoting and new-business entry consumes the most staff time at 82% of agencies, followed by policy renewals at 62%. Overall, 93% name one or the other, and no one chose “largely automated”. The mix shifts with size: certificates of insurance are named by 21% of small agencies and 53% of midsize, and claims intake by 2% of small and 27% of large. The open comments say the same thing in other words: 5 of 20 ask for AI help populating quotes, forms and carrier portals (paraphrased).

Exhibit 7

Quoting and renewals dominate the workload

Q7Which workflow consumes the most staff time today? (worded “most”; multiple selections collected)

Our reading. An agency that has not counted its hours cannot choose where AI pays back first.

1 respondent did not answer; 2.8 selections per respondent. Source: InsuredMine AI Survey, 2026.

More broken seams, more lost hours

What separates the agencies that lose the most time is not how they rate their data flow in general, but how many specific seams break. Agencies that name 3 or more of the five integration gaps are about twice as likely to lose 5 or more hours a week (65%) as agencies that name 0 or 1 (32%). A general data-flow rating did not separate agencies the same way: 42% of those rating it 1 or 2 lose 5 or more hours a week, against 48% of those rating it 4 or 5.

Exhibit 8

Agencies that name more broken seams lose more hours

Q5 × Q6Agencies losing 5 or more, and more than 10, hours a week to manual entry, by how many of the five gaps in Exhibit 5 they name.

Lose 5 or more hours a weekLose more than 10 hours a week

Interpretation. This is an association in small groups, not proof that fixing a seam recovers the hours. The practical step is to count the broken seams in one workflow.

* Small group; directional. Group sizes: 0 or 1 seams n = 31, 2 seams n = 19, 3 or more seams n = 20. The 70 agencies that answered Q5. About 15% in every group have not measured their hours. Source: InsuredMine AI Survey, 2026.

Chapter 4 of 4

Owners want proof on one workflow

Demand is broad, and it compounds

88% of agencies name at least one AI capability they expect to adopt in the next 12 months, and only 12% say AI is not a near-term priority, including 17% of small agencies and none of the large. Interest follows use: 97% of the agencies already beyond email and documents (38 of 39) plan to adopt more, against 70% of those that are not (30 of 43), so the distance between the two groups may widen. Client communications (51%), cross-sell prediction (46%) and conversation intelligence (43%) lead.

Exhibit 9

Interest is broad across capabilities

Q8Which AI capabilities is your agency most likely to adopt in the next 12 months? (multiple selections)

* Small samples; read directionally. “Beyond email and documents” means using follow-up, chatbots, workflow automation or predictive scoring (Exhibit 2). Source: InsuredMine AI Survey, 2026.

Most agencies prefer native integration, with a caution on the question

54% of agencies call native integration between their CRM and AMS a must, and 66% rate its importance 4 or 5. Midsize agencies feel it most, at an average of 4.73 out of 5 against 3.73 at small agencies. The question set native integration against middleware, which may have tilted answers toward native, so read it as strength of preference, not measured need. The preference is not simply a response to pain: agencies that rate their data flow 4 or 5 are more likely to call native a must than those that rate it 1 or 2.

Exhibit 10

54% call native integration a must

Q9How important is it that your CRM natively integrates with your AMS, versus requiring middleware? (1 = middleware acceptable, 5 = a must)

Rate it 5, a mustRate it 4 or 5

By data-flow rating (directional)

* Small samples. Average uses those answering; 3 respondents did not answer. Data-flow split is Q9 by Q4, groups of 26 and 23. Source: InsuredMine AI Survey, 2026.

Owners want to see one workflow work

The first ask is clarity on which day-to-day workflows AI actually improves (34%), followed by a low-commitment pilot (17%) and a live demo on a similar agency’s data (16%). Together, 67% chose clarity, a pilot or a demo, while ROI models and peer references draw 4% each. The ask is sharpest where use is shallowest: 44% of agencies that have not gone beyond email and documents ask for workflow clarity first, against 23% of those that have.

Exhibit 11

Workflow clarity is the first ask; ROI models barely register

Q10What would most increase your confidence in adopting an AI-powered CRM solution? (single choice)

Resistance. 4 of 20 open comments, all from small agencies, voiced skepticism (paraphrased): a preference to keep client relationships personal, worry that AI adds checking time, and doubts about where the technology leads the agency channel. A deployment plan has to make room for people who prefer not to use AI and count review time as part of the work.

9 respondents did not answer. “Nothing would change my mind” was not offered. Source: InsuredMine AI Survey, 2026.

What an AI-ready agency looks like in five years

Exhibit 12

82% want staff on relationships and advice, with AI on the back office

Q13In five years, what does an “AI-ready” AMS agency look like to you? (multiple selections)

2 respondents did not answer. Source: InsuredMine AI Survey, 2026.

Want to see one workflow work?

Renewal-risk and cross-sell prediction are live. See them on a book like yours, or start with a pilot.

See it on your book

Where you start

Where you start depends on where you’re stuck

The survey describes where agencies stand. The plans below are our judgment and have not been tested in the field. Midsize and large patterns are hypotheses.

What the survey shows

Readiness averages 2.60. 66% sit at the shallow end. Privacy is a concern for 43% and unclear priorities for 42%. 38% ask first for clarity on which workflows AI improves.

30 days

Log time, re-entry and corrections on one repeated task for 2 weeks. Write one page of data rules: what client information goes into which tool, and who approves a new one. Test one narrow step with a named reviewer, and keep it only if it saves time without adding checking.

Success

A before-and-after number for one workflow.

Questions to ask any vendor, including us

Where does our client data go, and is it used to train models?
Can you show it working on a real quote?
What will staff still re-key?

For every agency: name an owner, record a baseline, keep a person responsible for consequential decisions and count review time as part of the work.

Ask every vendor: Which of these capabilities are live at agencies our size, which are in pilot and which are planned?

* Small samples: midsize n = 15, large n = 11. Source: InsuredMine AI Survey, 2026.

Context

How this compares

Two industry studies cover related ground. The findings point the same way on direction, and this survey adds measures that we did not find in either.

	Big “I” ACT Tech Trends Report 2026	Vertafore 2026 outlook reports	This survey
Who answered	Big “I” member agencies in a national email survey. Respondent count not published.	1,300+ independent agency professionals, owners and non-owners.	82 InsuredMine customer agencies (53 small, 15 midsize, 11 large).
AI use today	About 3 in 10 use no AI tools. 8% report regular, strategic use.	About 1 in 3 actively using AI and 1 in 3 experimenting (mid-year report, via ACT).	84% use some AI. 52% stop at email and documents or have not started.
Outlook	68% likely to increase AI use within a year.	Almost 1 in 4 confident AI will transform everyday work in 2026.	88% name an AI capability to adopt within 12 months.
Top concern	Privacy and compliance, about 24% (one answer).	Not reported in the summaries we reviewed.	Privacy and compliance, 43% (several answers).
Data flow and manual hours	Not measured. Disconnected systems: about 16% as top challenge.	Not reported in the summaries we reviewed.	CRM–AMS data flow 2.86 of 5. 22% lose over 10 hours a week.

Where the findings agree

Privacy and compliance lead the concerns, and interest runs well ahead of depth of use.

Where this survey differs

A CRM vendor asking its own customers, so it can measure CRM–AMS data flow and manual hours by agency size.

What this survey did not ask

Written AI policies, output accuracy and staff training. ACT found that over half of agencies have no written AI policy.

Not directly comparable: ACT’s 8% is regular, strategic use, while our 84% counts any use, including email drafting. ACT allowed one top concern; we allowed several. Our respondents are customers who chose to answer, so they may be more engaged than the market. Sources: Big “I” Agents Council for Technology, 2026 Tech Trends Report (Feb 2026); Vertafore, 2026 Agency Trends Outlook (Feb 2026) and 2026 Agency Technology and Growth Outlook (as cited by ACT, IA Magazine, 15 Jul 2026). We reviewed ACT’s report and published summaries of Vertafore’s reports.

Read this before you quote us

Methodology and limits

Who answered

About 1,000 InsuredMine customer agencies were invited. The main contact, usually an owner, principal or senior executive, answered 13 questions between mid-July and mid-September 2026. 82 final responses were included, roughly 8%. By size: 53 small, 15 midsize, 11 large and 3 unreported. 51% of respondents run agencies of 10 or fewer employees.

How to read the figures

Shares use the full base of 82, and every exhibit shows non-responses. Averages use those answering. Q3 and Q7 were worded as “biggest” and “most” but collected as multiple selection, at 1.7 and 2.8 picks per respondent. We report what respondents named. Size bands combine 7 reported bands to keep groups readable.

What it cannot tell you

Respondents are customers of one vendor who chose to answer. They may be more engaged with AI, or with us, than agencies that did not. Results describe this sample and are not market estimates.
The midsize (n = 15) and large (n = 11) groups move with a single response.
The data does not establish that weak data flow causes shallow adoption.
The native-integration question may have favored native answers, and Q10 offered no “nothing would change my mind” option.
Readiness and data-flow scores are self-reported. The hours and dollar figures are modelled from time bands and assumptions.
The plans are judgment. Open comments are paraphrased and anonymous.

Where this leaves us

Agencies adopted the AI that asked the least of them. The work that takes their week asks more, because it depends on data that is accurate and in the right place.

Pick one repeated piece of that work, a quote or a renewal.

Count what it costs.

Make the data under it trustworthy.

Expand when the result holds.

Which of these findings describes your agency least well? Tell us.

Start with one workflow.

Bring one workflow. We can look together at where the work happens, what your systems carry across, and what is worth measuring first.

insuredmine.com/contact-us

InsuredMine builds an AI-powered CRM for independent P&C agencies, designed to work alongside the agency management system. This note is separate from the survey findings. Source for all exhibits: InsuredMine AI Survey, 2026 (n = 82).