What to use instead of FactSet Workstation

The product

FactSet Research Systems is a Norwalk, Connecticut company whose year ends on 31 August, and the Workstation is its desktop. FactSet's Workstation page, read on 9 October 2026, describes it as one environment joining 800-plus data sources — FactSet's own datasets, third-party data and a firm's in-house content — keyed to permanent entity and security identifiers. Around that sit alerting, charting and screening; screening across 10 million-plus private companies for M&A work; portfolio performance, risk and attribution with 120-plus risk models and four optimisers; live-linked Excel and PowerPoint models; programmatic access to 1,100-plus datasets; and FactSet Intelligence, its AI layer for search, chat, transcript analysis and portfolio commentary.

FactSet's 10-K for the year to 31 August 2025 says who pays for it: 8,996 clients and 237,324 users, with buy-side firms — asset managers, asset owners, hedge funds, wealth managers — at about 82 per cent of organic subscription value and banks, brokers and private equity at 18 per cent. The fourth-quarter results filed on 30 September 2026 put annual subscription value at 2,565.2 million dollars, across more than 9,200 clients and 260,000 users.

Divide one by the other and you get a figure just under 10,000 dollars a user, which is not a price: the total includes data feeds, managed services and enterprise analytics alongside the desktops. FactSet's pricing page offers a personalised proposal and nothing else.

People leave the Workstation for the usual institutional reasons — a seat that ended with a job, a budget review, a fund too small to justify it — and they rarely used all of it. A portfolio manager, a banker, an advisor and a quant on the same contract use it four different ways.

What to use instead

Split it by who sat at the desk, because the Workstation is sold by firm type and each one used a different part.

The quant and the screener. Portfolio123 is the one pick that runs on FactSet's own fundamentals and estimates, point-in-time, with delisted stocks kept in the history. Ranking systems, screens and backtests over five to twenty years depend on the plan; US, Canada and Europe are covered, Asia is not. If you screened rather than ranked, GuruFocus has thirty years of financials and a screener with backtests to 2006 on its upper tier, on Morningstar fundamentals.

The analyst with a model. FactSet's live-linked Excel is the hardest piece to replace, and two picks cover it from opposite ends. Daloopa writes reported figures, segments, KPIs and guidance into the analyst's own workbook, each linked to the page it came from — sold by quote, to firms. Wisesheets is the self-serve version: worksheet functions that take ranges of tickers and periods, at 60 dollars a year on a personal licence. GuruFocus also sells an Excel add-in, capped at 2,000 queries below Premium Plus.

The advisor. FactSet sells the Workstation to wealth managers with client portals and proposal tools. YCharts is built for that seat — research, portfolio construction and branded client reports — and names FactSet among its financial-data suppliers. It is quote-only too, but unlike FactSet it publishes a plan aimed at individual investors.

The developer. FactSet's API is priced the way everything else there is, by proposal. Financial Modeling Prep sells statements, estimates, transcripts and 13F holdings over REST with published prices, on personal-use terms below its quoted Enterprise plan.

The transcripts. Quartr carries the calls themselves — live audio, real-time transcripts, decks and filings — free on mobile. For a transcript archive as data, the earnings-call transcripts how-to lists the routes.

What you give up

Portfolio analytics. Attribution, risk decomposition across 120-plus risk models and a choice of optimisers are a large part of what buy-side firms pay FactSet for, and nothing in this catalogue sells them as a service. The analysis libraries — Riskfolio-Lib, PyPortfolioOpt — do the mathematics in Python, on data and factor definitions you supply yourself.

One identifier under everything. FactSet keys every dataset, its own and a client's, to permanent entity and security identifiers, so a holding, a filing and an estimate join without effort. With four vendors you join on tickers and ISINs and maintain the mapping; CUSIP, ISIN and FIGI covers what each identifier survives.

The consensus you quoted. A consensus figure belongs to whoever collected it. The picks carry other suppliers' — GuruFocus names Refinitiv and Morningstar, the S&P resellers carry S&P's — so an EPS consensus quoted from FactSet in a past memo will not reappear unchanged. Different contributors, different cut-off times.

Migration notes

Keep the data licence if the data was the point. Portfolio123 lets customers who already license FactSet download point-in-time data through its API and screener reports; without one, a recent screen downloads values for the first 500 tickers only and an older one downloads blank. A firm dropping Workstation seats but keeping a FactSet data feed has a cheaper configuration available than starting over.

Convert live links before access lapses. Every cell in a FactSet-linked workbook is a request to FactSet. Save a values-only copy of each model you need to keep, and note the field codes beside it, so the rebuild in Daloopa, Wisesheets or a script knows what each cell meant.

Expect point-in-time to move. Portfolio123's own documentation says FactSet backfills figures released after an announcement, so a backtest re-run later can change. That was true inside the Workstation as well; it is simply more visible when you run the backtest yourself.

Read the reseller's terms, not FactSet's. A screen on FactSet data inside Portfolio123 or YCharts is governed by that vendor's licence, and personal-use plans such as Wisesheets' and Financial Modeling Prep's do not cover a number circulated round a team. Anything shown to a client is redistribution.