Free toolsState of the Referral 2026
When a client passes along your name, the friend looks you up that night. We read 985 independent advisor websites the way that friend would, one step at a time, and counted where the trail holds and where it goes quiet. These are RIA website benchmarks you can hold your own site against.
Fig. 1A composite homepage, not a real firm. 21% of the sites we read say "peace of mind." October 2026.
The short version
8 of 985
kept the line at all seven steps a referred prospect takes. The median site passes 4 of 7.
7.2%
post their fees with a number. Of the firms that call themselves fee-only, 85% never say what the fee is.
28%
say on the homepage what kind of client they serve. The rest are for "individuals and families."
13%
let a visitor book a call online. 38% offer neither a booking link nor a number you can tap.
The walk
Each step is something the friend checks before they call. The big number is the share of sites that pass that step. The brass number is the share still on the line after every step so far.
Of 985 sites, 8 pass all seven steps (0.8%). Who you serve and fees are where the line goes quiet.
Search
66% of sites have a page title and a written description for the search result. 30% have no description at all, so Google writes one from whatever text it finds first, which is often a menu or a disclosure. 24% have no main headline (H1) on the homepage.
The fix Write a 150-character description that says who you serve and where.
Homepage
94% load over HTTPS and fit a phone screen. This is the step most firms pass. The median homepage was 112 KB of HTML and answered our crawler in 0.36 seconds.
The fix If yours is not in the 94%, that is the first fix.
Who you serve
28% name a specific kind of client on the homepage. The most common is business owners (15% of sites), then retirees and pre-retirees (7.4%) and executives (5.2%). The rest say some version of "individuals and families," which is everyone, which is no one in particular.
The fix Say who you are for in the first two lines, in their words.
Fees
7.2% post a fee or fee schedule with a number. 27% of sites say "fee-only," and of those, 15% say what the fee is. The number is already public in Form ADV Part 2A; the site just makes the friend go and dig.
The fix Post the schedule, or at least a starting point and a minimum.
Team and bios
87% have a team or about page with bios. This is the bright spot. The median team page has about 9 bios, and 50% of sites mention the CFP mark. The bios do read alike: "outside of work" appears on 16% of sites.
The fix One specific sentence per bio about the clients that person serves.
The SEC check
83% link to Form CRS or Form ADV. We could not find a Form CRS link on 25% of sites (homepage plus up to four pages), and 17% link to neither. 20% link straight to the SEC's adviser search, which is where a careful friend goes anyway.
The fix Link Form CRS and your IAPD page in the footer of every page.
Booking
13% let a visitor book a call online (Calendly is 69% of those). 57% have a tap-to-call phone link, 28% show a number you cannot tap, and 38% have neither a booking link nor a tap-to-call number. At 9:12 p.m. that leaves a contact form.
The fix Add a booking link next to the phone number, and make the number tappable.
This is a plain-language read of public websites, not legal or compliance advice. Your compliance reviewer has the final word on what goes on your site.
By firm size
The same seven steps, split by regulatory AUM. Firms over $1B link to Form CRS more often (82% against 72%), and smaller firms are a little more likely to offer online booking. Everywhere else, the gaps are a few points.
Steps passed per site
Only 0.8% pass all seven. 13% pass two or fewer.
The phrase leaderboard
Ranked by the share of sites that use them on the homepage or the about and team pages. 87% of sites use at least one. The median site uses 3. One site used 16.
Counted once per site, with close variants grouped ("your best interest" and "your best interests"). Left out: legal boilerplate, navigation, job titles, degrees, and the word "fiduciary," which 53% of sites use and which is a legal standard, not a cliché.
Smaller things a referral notices
have a footer copyright year that is two or more years old.
use stock photography we could spot by file name or subject: beach 3.5%, happy retired couple 2.2%, compass 2.1%, lighthouse 1.7%, handshake 1.6%, sailboat 1.5%.
show testimonials or reviews, which the SEC Marketing Rule has allowed, with disclosures, since November 2022.
have a video on the homepage.
show a phone number on the homepage that a phone cannot tap to call.
of the sites we reached run on WordPress. Then Squarespace (10%) and Wix (6.9%).
The full report
The walk, the by-size tables for all 13 measures, the phrase leaderboard, and the method, laid out to print or forward to a partner. We will send a copy to your inbox too.
Unlocked
Or save this page with your browser's print command. The appendix tables are below.
| Measure | All | $100M to $500M | $500M to $1B | $1B to $2B |
|---|---|---|---|---|
| Post their fees or fee schedule on the site | 7.2% | 7.9% | 6.1% | 7.2% |
| Name a specific kind of client on the homepage | 28% | 26% | 31% | 29% |
| Link to Form CRS | 75% | 72% | 75% | 82% |
| Link to Form ADV or adviserinfo.sec.gov | 56% | 53% | 57% | 62% |
| Have a team or about page with bios | 87% | 82% | 92% | 91% |
| Let a visitor book a call online | 13% | 15% | 11% | 11% |
| Tap-to-call phone link | 57% | 55% | 58% | 60% |
| Show testimonials or reviews | 14% | 14% | 15% | 12% |
| Have a video on the homepage | 14% | 12% | 17% | 15% |
| Use stock-photo tropes (sailboats, beaches, handshakes) | 25% | 24% | 28% | 23% |
| Footer copyright year is 2 or more years old | 11% | 12% | 11% | 8.6% |
| Mobile-ready | 96% | 95% | 97% | 97% |
| Served over HTTPS | 96% | 95% | 97% | 96% |
| Sites read | 985 | 481 | 295 | 209 |
| Why we could not read a site | Sites |
|---|---|
| Blocked our crawler (403/401/429) | 164 |
| Page built by JavaScript (text not in HTML) | 56 |
| Domain did not resolve or refused | 10 |
| Certificate error | 9 |
| Listed URL returned 404 | 7 |
| Server error | 7 |
| Other (robots.txt, timeouts, odd server replies) | 12 |
| Total not read | 265 |
Check your own firm
The same checks, run on your homepage, in about 30 seconds. No login.
Method
The firms. We started from the SEC's Investment Adviser Information Report for SEC-registered advisers (data as of September 1, 2026, posted October 2026), 17,210 firms. We kept firms based in the US with $100M to $2B in regulatory assets under management, where individual and high-net-worth clients account for at least half of those assets, with no private funds and no registered funds, 150 or fewer employees, and a firm website on file (not a social profile). That left 3,578 firms. We drew a stratified random sample of 1,250: 50% from $100M to $500M, 30% from $500M to $1B, and 20% from $1B to $2B. The median firm in the sample manages $499M and has 6 employees.
The walk. On October 24, 2026 we fetched each homepage and up to four pages linked from it (about or team, fees, services, contact), eight sites at a time, with a 10-second timeout, a crawler name and contact email in every request, and robots.txt respected. We reached 1,041 sites and could read 985. The rest blocked automated visits, were built entirely in JavaScript, or did not load (Appendix B in the full report).
The steps. Search: a page title and a meta description of 50 characters or more. Homepage: served over HTTPS with a mobile viewport. Who you serve: the homepage names a kind of client beyond "individuals and families" (an occupation, a life stage, a situation). Fees: the site states its own fee with a number. Bios: a team or about page with bios. The SEC check: a link to Form CRS, Form ADV, or the SEC's adviser search. Booking: a link to an online scheduler. "Still on the line" counts sites that pass every step up to that point.
Checks. Signals come from the same open rules the Referral Walk uses (signals.js version 1.1.0), cross-checked against a second, independent extractor. We read every fee result by hand: 61 of 71 automatic hits were a firm stating its own fee (86%), and the hand count across both extractors found 7.8% of sites post fees, close to the 7.2% shown here.
Limits. We read page HTML, not what a browser paints, so content loaded by scripts and anything in PDFs or past four clicks is missed; some steps will read a little low. Search is judged from the page's own title and description, not live rankings. AUM comes from each firm's last Form ADV filing and can be up to a year old. Firms that block automated visits are not in the numbers. We do not name, score, or rank any firm, and we keep no copy of any site beyond this analysis.
Cite this
Furlong. "State of the Referral 2026: what 985 RIA websites show a referred prospect." October 2026. https://furlongco.com/tools/state-of-the-referral/
Questions
We started from the SEC's Investment Adviser Information Report (data as of September 1, 2026). We kept SEC-registered firms with $100M to $2B in regulatory assets where individual and high-net-worth clients are at least half of assets, no private funds, and a website on file. We drew a stratified random sample of 1,250 (50% $100M to $500M, 30% $500M to $1B, 20% $1B to $2B), reached 1,041, and could read 985.
No. Every number is an aggregate. We do not publish a list of firms, scores for named firms, or any site we read.
The site states its own fee with a number: a percentage, a flat fee, a retainer, an hourly rate, or a schedule. Saying "fee-only" or "competitive fees" without a number does not count. We hand-checked the fee results; see the method.
We could only read what is in the page HTML, on the homepage and up to four linked pages. A booking link buried on page six, or a fee schedule in a PDF, would be missed. So treat each number as "a referral who clicks around a little would find this," not a census.
Yes. The Referral Walk runs these checks on any homepage in about 30 seconds, and Roast My Advisor Bio reads a bio for the phrases on the leaderboard.
Yes, with a link. Cite it as shown in the "Cite this" box. The data is aggregate and the method is on this page.