In June 2024, Equilar launched ERIC — the Equilar Research Intelligence Copilot. It is an AI-powered tool that reads proxy statements, 10-Ks, and 8-Ks, lets you filter by peer group and industry, runs year-over-year comparisons, and cites its sources so the answers don’t hallucinate. It sits on top of the 20-year proprietary dataset Equilar has been building since 2000.
That is, by any reasonable definition, a serious AI analyst for executive compensation work. So a question we get asked often is the obvious one: what does Spencer do that ERIC doesn’t?
The honest answer isn’t “Spencer has better data” or “Spencer’s AI is smarter.” Both products are built on top of the same public SEC corpus, and both will get better at the same pace as the underlying models improve. The real answer is about which side of the consulting engagement each product sits on — and why a boutique comp firm will end up using both.
§01Two AI analysts, two different jobs
Read ERIC’s marketing carefully and the customer profile is consistent. The case studies are about CHROs prepping for board meetings. The featured use cases are “stay ahead of shareholder feedback” and “summarize peer-group voting outcomes before our annual proxy.” The integrations are with Equilar’s CHRO Navigator and Insight platforms — which serve corporate compensation teams at filers.
ERIC, in other words, is built for the side of the engagement that files the proxy. The corporation. The CHRO and her team. The comp committee preparing to approve a CEO’s package and defend it to shareholders.
Spencer is built for the other chair at the table. The boutique executive compensation consulting firm advising that corporation. The senior partner who structures the CEO offer, benchmarks against peers, writes the recommendation, and presents to the comp committee. Different reader. Different deliverable. Different accountability.
Once you see the side-of-the-table distinction, a lot of things about both products fall into place.
§02What “serves the filer” actually means
The filer’s questions are inward-facing. What did we disclose last year? How does our pay design compare to our peers? What will ISS say about it? What are our shareholders likely to push back on? These are the questions ERIC is designed to answer, and answer well.
Because the audience is one company at a time, the workflow is broadly generic. ERIC ships the same analytical surface to every corporate user. That’s appropriate — a CHRO doesn’t need ERIC to learn her house style, because her house style is whatever she decides it is. There’s no template library to match. There’s no second client whose conventions are different.
The output goes to her own comp committee, in her own format. ERIC is the research desk. The deliverable is internal.
§03What “serves the firm” looks like
A comp consultant’s job is structurally different. A 12-person boutique might serve forty corporate clients across a single annual cycle. Every one of those forty engagements produces a deck that gets delivered to a comp committee with the firm’s name on it.
The deck has to look like the firm’s deck. Not Equilar’s. Not ERIC’s. Not a generic AI tool’s. The firm has a peer-group methodology it has spent two decades refining. It has percentile cuts it always uses (25/50/75 or 10/25/50/75/90, depending on the firm). It has a committee-naming convention. It has a way of writing the CD&A summary that senior partners have agreed on. It has a house philosophy on whether equity vesting should be cliff or graded for early-stage biotech CEOs.
Those things are the product. They are why a compensation committee hires this firm instead of another. An AI tool that ignores them — that produces generic, industry-standard outputs — isn’t a labor multiplier for that firm. It’s a competing tool the partner has to spend time translating intothe firm’s shape before it’s usable.
Spencer’s entire architecture is built around the opposite premise. The product surface looks like the firm’s process, not a generic data browser. Templates, peer-group memory by client, custom column orders, custom percentile cuts, firm-specific committee naming, house-style CD&A drafting — all of that is treated as a feature, not a customization burden.
ERIC is for the corporation deciding what to pay its CEO. Spencer is for the firm advising the corporation deciding what to pay its CEO. Both can be excellent at their jobs. They’re different jobs.
§04Why the boutique firm will end up paying for both
We get the procurement question a lot: if we already have an Equilar subscription, why would we buy Spencer too? The right way to think about that is as two different categories of spend.
Equilar is a data subscription. It sits on the same budget line as Bloomberg or FactSet — the cost of having the underlying corpus available to your firm. ERIC is part of that subscription for many tiers. It’s a research-desk AI on top of the data, and it’s genuinely useful for the moments a consultant needs to be in the corporate seat — preparing a client for a board meeting, simulating how ISS will read a disclosure, scanning shareholder feedback across a peer group.
Spencer is a labor subscription. It sits on the line item where you’d hire another analyst. The decision is not “data or no data” — you already pay for data. The decision is “do we hire a fifth analyst this year or do we let Spencer be the analyst.” The math at a 12-person boutique is usually pretty fast.
Equilar / ERIC — the corporate side’s research desk. Sits next to Bloomberg in your budget.
Spencer — the firm’s analyst bench. Sits next to your headcount line in your budget.
§05The deeper question: what happens when clients have AI too?
There’s a quieter anxiety running through the boutique comp community that doesn’t come up on sales calls but absolutely comes up over drinks at the WorldatWork conference. It goes something like: if my clients can use ERIC to do their own peer benchmarking, do they still need me?
This is the right question to be asking, and the answer is yes — but only if the firm’s answer to that question is sharper than “we’ve always done it this way.”
ERIC threatens junior-analyst hours at corporations. It accelerates the work a CHRO’s internal team would have done before engaging an outside consultant. That’s real. But it does not replace the judgment a senior partner brings: the structured recommendation, the narrative for the committee, the read on whether a CEO’s package will pass say-on-pay, the institutional context across dozens of similar engagements.
The firms that will thrive in the next ten years are the ones whose partners spend more of their time on judgment and less of their time on Ctrl-F’ing through 200-page proxies. The firms that will struggle are the ones that try to compete with their clients’ AI tools by doing the same work, only slower.
Spencer’s job in that future is to absorb the structure work — the peer-group construction, the disclosure extraction, the year-over-year diffs, the deck assembly — so the consultant spends her hours on the part of the work clients can’t do with ERIC alone: what should we recommend, and how do we tell the story.
§06Three principles that follow from the side-of-the-table framing
Once you accept that Spencer is the consulting firm’s analyst, a few product decisions stop being negotiable.
Show the work, always
Senior partners stake their reputations on the numbers in their decks. An AI assistant that produces unverifiable claims is worse than no assistant, because every output has to be re-verified by hand before it’s signable. Every number Spencer produces is one click from its source disclosure — not as a feature, but as a precondition for the product existing at all.
Confidently uncertain
The most insidious failure mode for any AI in this domain is filling a data gap with a plausible-sounding inference and presenting it as fact. A “not disclosed” beats a confident wrong number every time. Confident wrong answers are how a firm loses a client forever. We’d rather Spencer say it doesn’t know than guess.
Your firm’s structure, not the industry’s
ERIC ships the same surface to every corporate user, because the corporate user is asking inward-facing questions about their own company. A consulting firm is asking the same questions across forty clients, in a shape that has to look like the firm’s. So Spencer learns each firm’s templates, peer-group conventions, and house style — not as a polish item, but as the central feature that makes the product worth paying for.
§07Where we go from here
We don’t view ERIC as a competitor. We view it as evidence that the category we care about — AI for executive compensation work — is now real enough that the market-leading data vendor has built into it. That validates the thesis. It also clarifies the seat Spencer needs to take: the consulting firm’s seat, not the corporate filer’s.
Most boutique comp firms we talk to already pay for Equilar. Several already use ERIC for the parts of their work where they need to be in their client’s headspace. Spencer is what they reach for the rest of the time — when they need to operate as their firm, producing their firm’s output, at a fraction of the analyst-hour cost they used to spend.
If you run a boutique comp practice and any of this resonates, we’d like to talk. The fastest way to see what Spencer does is to sign up and run a board demographics report against a peer set you know well — then judge for yourself whether what comes out the other end looks like your firm or looks like everyone else’s.