Question 1

What is ValuationMCP, in one sentence?

Accepted Answer

It is a valuation engine for wealth-management practices that your AI assistant can use directly: it gathers the figures that already live in your systems, benchmarks them against SEC Form ADV filings and published industry data, and returns a seven-method valuation with its range and a data-confidence score.

Question 2

What is MCP, and do I need to understand it?

Accepted Answer

No. The Model Context Protocol is the open standard that lets an AI assistant such as Claude use outside tools. You connect ValuationMCP once, then ask your assistant in plain language: "Value my practice", "What should I not pay more than for this firm?", "What happens if I step away?"

Question 3

Which AI assistants work with it?

Accepted Answer

Any assistant that supports MCP. Claude is the reference client. Enterprise agents can also reach it over A2A — see the Enterprise section.

Question 4

Do I need an account?

Accepted Answer

The free deal-room example needs no account. Everything else does: every tool runs as you, over your own practice data, behind OAuth. Creating an account takes a minute.

Question 5

What does it cost?

Accepted Answer

Your first full valuation is free. Owner is $39/mo or $390/yr (founding members $290/yr while the first 100 last), Deal is $149/mo or $1,490/yr, and Network starts at $20 per seat per month. See pricing.

Question 6

Is this a formal appraisal?

Accepted Answer

No. ValuationMCP produces a modeled estimate for planning discussions, always shown with its range, its data coverage and the assumptions behind it. It is not an appraisal, a fairness opinion, an offer, or investment, legal, tax or M&A advice. When a situation has outgrown a modeled estimate — a dispute, a divorce, an estate — the output says so and points you to a credentialed appraiser.

Question 7

How accurate is the number?

Accepted Answer

As accurate as the inputs and the published benchmarks allow, and the engine tells you which. Each run reports how complete your inputs are (14 facts drive the math), which method ran on sourced figures versus a placeholder, and how far apart the seven methods landed. A number built on 14 of 14 sourced inputs is a different thing from a first screen built on 8, and the report never lets the two look alike.

Question 8

What are the seven methods?

Accepted Answer

Revenue multiple, EBITDA multiple, EBOC multiple (profit before you pay yourself), discounted cash flow, client value, AUM percentage, and comparable transactions. They are blended into one fair estimate, each weighted by the confidence of the data behind it, then adjusted by a six-factor book-quality score. Ask your assistant for "Valuation Literacy" if you want the mechanism explained in full.

Question 9

Where do the benchmarks come from?

Accepted Answer

Three named sources, each with one job. Transaction multiples come from the published Sica Fletcher SCF Index for your size tier (watched weekly; a new edition is reviewed before it is applied). Firm size and client-mix percentiles are computed from 23,740 SEC Form ADV filings. Operating medians — revenue per client, margin, productivity, attrition — come from the InvestmentNews Advisor Benchmarking Study. Other publishers' multiples (Echelon, CT Acquisitions, Succession Resource Group) are shown alongside for comparison and are never an input. Full detail is on the homepage under "Where the numbers come from".

Question 10

What happens to my data?

Accepted Answer

Only the provenance of each figure (which platform supplied which field, and when) and the calculated outputs are stored. No client lists, no account-level data, no documents. Every pull from your systems happens inside a conversation you can see, as a one-time request; nothing keeps a standing key to your CRM or ledger. See the Privacy page for the full policy.

Question 11

Can I just type my numbers in?

Accepted Answer

Yes. Your assistant will ask for what the math needs, and you can state figures directly instead of connecting a platform. Stated figures are labeled as such, and if a stated figure disagrees with one a system supplied, yours is used and both are kept with their source.

Question 12

Can I use it in a browser instead of an assistant?

Accepted Answer

Yes. Every saved run opens in the workbench at valuationmcp.ai, where the same numbers are editable with charts. The assistant is the fastest way to gather inputs; the workbench is where you adjust them.

Question 13

What do I get on the sell side?

Accepted Answer

An asking price (the usual listing premium included), the fair estimate underneath it, the range the seven methods landed in, your book-quality score with each factor in dollars, where you sit among filed RIAs your size, and the market you would be selling into. Every figure carries its source.

Question 14

Which of my systems can it pull from?

Accepted Answer

Any of the 52 supported MCP servers across 8 source categories — CRM, portfolio accounting, custodian, billing, ledger, planning and more. See the full list at /mcp-servers. If your platform is not there, you can state the figures and the report labels them as stated by the owner.

Question 15

What drives the number up or down?

Accepted Answer

Six book-quality factors — recurring revenue, client concentration, client demographics, growth, succession readiness, and operating margin against firms your size — plus the published multiples for your tier. Ask "What's holding my score back?" for each factor's dollar effect, and "How do I move the number?" for where you sit against every published value-driver anchor. The engine prices the arithmetic; the decisions stay with you and your advisors.

Question 16

My number came out lower than I expected. Why?

Accepted Answer

Usually one of three things: a placeholder multiple where a sourced one was missing, an operating margin below what firms your size report, or a gap in the inputs that the quality factors are silently assuming conservatively. Ask "What's missing, and is it worth getting?" — it turns your confidence score into a to-do list ranked by how many dollars each missing fact is costing.

Question 17

Can I check my figures before I rely on them?

Accepted Answer

Yes. "Do these numbers hold up?" checks your inputs against your own filings and published medians, names each disagreement, and tells you where to go settle it. It raises questions for you; it never corrects the model on its own.

Question 18

How do I prepare for a buyer's questions?

Accepted Answer

The deal room. From one saved run, "Deal room rehearsal" argues the buyer's objections, drafts your answers, lays out likely terms and gives you a verification checklist. Your prep stays private. You can also host a shared room and invite a buyer; each side's preparation stays on its own side. Try the free example.

Question 19

What happens when a buyer sends me a data request?

Accepted Answer

Requests reach your firm's inbox only after CRD verification and your opt-in, and can be gated behind an NDA. Your assistant previews the figures and their sources, and nothing is shared until you approve it. Your own valuation is never part of what is sent.

Question 20

How often should I re-run it?

Accepted Answer

Whenever something material changes — a quarter closes, a large client leaves, a successor is named. "What changed since last time?" compares two saved runs and attributes the movement to business figures, data completeness, assumptions or settings, so a change in the model is never mistaken for a change in the market.

Question 21

What do I get on the buy side?

Accepted Answer

A walk-away ceiling for a target — the most the economics support, not an offer — with its range, its data coverage, what diligence would need to replace, and the market you are buying into. The same seven methods run; the headline flips.

Question 22

How do I start with a firm I'm looking at?

Accepted Answer

Give your assistant the target's SEC CRD. Its filed assets and client count fill in from the Form ADV via advmcp.ai; anything the target has shared with you is added as target-provided and always wins over a filed or assumed figure.

Question 23

Can it find targets for me?

Accepted Answer

With advmcp.ai connected, yes. "Find and screen acquisition targets" turns your acquisition thesis into a ranked shortlist from filed Form ADV data and puts a screening ceiling on each firm. It is a research list, never an approach list: nothing in a public filing says a firm wants to sell.

Question 24

How can it value a firm it has no financials for?

Accepted Answer

It can't, honestly, and it says so. Form ADV carries assets and client counts but no revenue, profit, margin or retention. A screening ceiling is built on cited industry medians and labeled low-confidence by construction. "What should I ask for before I buy?" lists what to request so a real valuation can replace the screen.

Question 25

How do I compare several targets?

Accepted Answer

"Which of these is the better buy?" reads two or more targets side by side — fair estimate, ceiling, book quality, concentration, continuity, data confidence — with the tradeoffs named and no winner picked. "What does this add to my book?" reads a target against your own practice: the combined size tier, how concentration and client mix shift, and where the two books overlap.

Question 26

Will the seller see my ceiling?

Accepted Answer

No. Your walk-away ceiling is never shared, including inside a shared deal room. Buyer prep argues the seller's case back at you privately.

Question 27

Can I ask a target for data through it?

Accepted Answer

Yes. "Request data from a target" prepares a private, voluntary request with a dry-run preview; the email is left as a draft for you to send, and the target answers from its own records only after its owner approves. Requests can be put behind an NDA.

Question 28

Does a target know I'm looking?

Accepted Answer

Not unless you contact them. Screening and ceilings run entirely on filed public data and your own inputs. A firm only hears from you when you choose to send a request.

Question 29

Who uses this at an enterprise level?

Accepted Answer

Broker-dealers, RIA aggregators, custodians and platforms that want every affiliated advisor to have a defensible, consistent view of what their practice is worth — and what would move it — without staffing a valuation desk.

Question 30

How does an enterprise agent connect?

Accepted Answer

ValuationMCP is MCP-native and publishes an A2A Agent Card, so an enterprise agent running on Bedrock AgentCore, Azure AI Foundry or Agentforce can send or answer requests through the same owner-approved flow as a human user's assistant. Details are at /agent-to-agent and /developers.

Question 31

Can we run it across many advisors?

Accepted Answer

Yes — each advisor authenticates as themselves, so the data stays segmented by owner and every tool runs over that advisor's own practice. Firm-level claims verify a firm's CRD so incoming requests reach the right inbox. Today, contact us at hnycomb.ai to discuss a rollout.

Question 32

What does "practice management" mean here?

Accepted Answer

Using the valuation as a management instrument, not just a deal number. The engine prices six book-quality factors in dollars, places the practice against state and size cohorts from filed RIAs, and shows where it sits against every published value-driver anchor. "What's holding my score back?", "How do I move the number?" and "How do I compare to firms like mine?" are the workflows; running them quarterly turns the valuation into a scorecard.

Question 33

What is the succession-readiness offering?

Accepted Answer

Four workflows that read the continuity picture already inside a valuation. "What happens if you step away?" reads who owns the relationships, what is on paper and how much runway there is. "What to put on paper before you hand it over" prices the document-level moves — a named successor, a written plan, a funded buy-sell — each with the engine's dollar effect, because paperwork is the cheapest lever an owner has. "What does this number mean for an internal buyout?" reads the estimate for a G2 or partner buy-in: what tends to be financeable versus seller-carried. "Does your buy-sell still cover the number?" compares an existing agreement's funded amount to current value and names the gap in dollars. See them all on /skills.

Question 34

Does it give advice on deal structure, financing or insurance?

Accepted Answer

No. It describes what the number implies and routes the structural questions to your attorney, CPA and insurance advisor. "Argue it from every side" puts the same computed number in front of the seller's broker, a buyer, a lender, a practice manager, succession planning, a business partner, the CPA's after-tax lens and the bar a credentialed appraiser would set — each a lens, never a character with invented figures.

Question 35

Can advisors see each other's valuations?

Accepted Answer

No. Access is per owner, behind OAuth. Sharing happens only through a deal room an owner opens or a data request an owner approves.

Question 36

What are "artifacts"?

Accepted Answer

Ten of the workflows also produce a page you can reopen — a benchmark board, a council memo, a succession read — built only from figures the tools returned and updated in place by a later run. They carry their own provenance footer (engine version, index version, data confidence, date), so a page that gets forwarded still says where its numbers came from.

Question 37

Can we white-label or embed it?

Accepted Answer

Not today. Enterprise agents can call it over A2A under your own agent's name, and the workbench and artifacts carry ValuationMCP branding. Talk to us at hnycomb.ai about what you need.

Question 38

How current is the data, and who maintains it?

Accepted Answer

The SCF Index source page is watched weekly and a new edition is staged for review rather than applied automatically. Form ADV percentiles are rebuilt from the SEC's IAPD compilation feed (the current feed date is shown on the homepage). InvestmentNews medians are updated with each annual study. Every run records the index version it priced off, so two runs are always comparable.

Question 39

Where do I report a problem or ask something that isn't here?

Accepted Answer

hnycomb.ai — ValuationMCP is part of the HnyComb network alongside WealthMCP, ADVmcp and 5500mcp.

Question 40

What's included in the free tier?

Accepted Answer

One practice, one full valuation with every method, the range and the data-confidence score, the Practice valuation run and Valuation literacy skills, and a one-page artifact. The workbench and history are view-only.

Question 41

What happens on my second run?

Accepted Answer

On the free plan, reruns return a preview: the fair estimate, range and confidence. The method breakdown, quality table, sensitivity, peers and next steps unlock on Owner. Your assistant says so up front on the first run, so there is no surprise mid-conversation.

Question 42

Can I switch monthly to annual?

Accepted Answer

Yes, any time in the billing portal from your account's Billing page. Annual is two months free compared with paying monthly.

Question 43

What does "1 sandbox practice" mean?

Accepted Answer

Owner covers your own firm plus one more practice you can use to test what-ifs, such as a merger partner or a successor's book, without touching your real record.

Question 44

Do you meter API calls?

Accepted Answer

No. There is no per-call metering on any paid plan: rerun as often as you like. A fair-use rate limit protects the service from runaway automation.

Question 45

What's the refund policy?

Accepted Answer

Annual plans have a 30-day, no-questions refund. You can cancel any plan in the billing portal; it stays active until the end of the period you paid for. Details are in the Terms.

Frequently asked questions