AI guides
How the models compare, what their features actually do, and how to get answers you can check. Every guide lists its sources and the date its facts were last checked.
Model comparisons
Claude Haiku vs Sonnet vs Opus
Claude Haiku, Sonnet and Opus 5.5 compared: price, speed, which plans include each, where Fable fits above them, and a quick test to pick the right model.
Claude Pro vs ChatGPT Plus
Claude Pro vs ChatGPT Plus, line by line: models, usage limits, Research and deep research, files, what you lose with each, and when Max or Pro is worth it.
Claude vs ChatGPT
Claude vs ChatGPT on what matters at work: which is more accurate, what each plan costs, which features you get, and how to test both on your own questions.
Claude vs ChatGPT for data
Claude vs ChatGPT for data analysis: file limits, code execution, what the benchmarks show, where metric definitions trip both up, and a test to run on each.
Claude vs ChatGPT for research
Claude Research vs ChatGPT deep research: how each works, what the citation studies tested (and didn't), which to use for which job, and how to check a report.
Claude vs ChatGPT vs Gemini
Claude, ChatGPT and Gemini compared on accuracy, plans and everyday work after their autumn 2026 launches, with the scoreboards that disagree and a test to run.
Claude vs Gemini
Claude vs Gemini after Opus 5.5 and Gemini 4 Argon: which gets more right, where they tie, what each plan costs and how to test both on your own documents.
Claude vs Perplexity
Claude vs Perplexity: an assistant for your own documents and data against a search engine that cites everything. Plans, models, accuracy evidence and a test.
Gemini vs ChatGPT
Google Gemini vs ChatGPT after Gemini 4 Argon: what each plan costs, which is more accurate on the independent indexes, and a quick test to run on your own work.
Perplexity Pro vs ChatGPT Plus
Perplexity Pro vs ChatGPT Plus: what $17 and $20 a month buy in models, limits, research and tools, why paying more didn't buy accuracy, and how to test both.
Perplexity vs ChatGPT
Perplexity vs ChatGPT: an answer engine that cites everything against an assistant that does the work. Accuracy studies, plans, models and a two-minute check.
Claude features
Claude artifacts
What Claude artifacts are, how to make, edit, export and share them, which plans get which types, and a two-minute check before you send one on.
Claude connectors
How Claude connectors and MCP work, how to add one, what each plan allows, the security risks Anthropic warns about, and how to check what a connector returned.
Claude Dashboards
What Claude Dashboards does, which plans have it, how to build one from a warehouse connector, how it refreshes, and a two-minute check on every chart's query.
Claude data analysis
How Claude analyses CSV and Excel files with code execution, its file limits, where its analysis goes wrong (definitions, filters, joins), and a quick check.
Claude for Excel
How to install Claude for Excel, what it does on Pro, Max, Team and Enterprise, its documented limits, seven prompts with checks, and how to verify its work.
Claude memory
How Claude's memory saves context between chats, how to view, edit, pause or reset it, what it never stores, and how to spot an answer built on a stale memory.
Claude Projects
What Claude Projects are, how to write project instructions and pick knowledge files, plan limits and sharing, and how to check Claude used your files.
Claude Research
How Claude's Research mode works, which plans have it, how long it takes, how to brief it, and a two-minute check for the citations in its reports.
Claude skills
What Claude skills (Agent Skills) are, how to install or write one, which skills are worth having for analysis work, the security risks, and how to test one.
Prompting for reliable answers
Context engineering
Context engineering explained: what it covers beyond the prompt, why more context can make answers worse, and four habits that keep long AI tasks accurate.
Data analysis prompts
Five copy-paste prompts for data analysis that make ChatGPT or Claude state its definitions, calculate in code, flag missing data and keep assumptions apart.
Financial analysis prompts
Copy-paste ChatGPT and Claude prompts for financial analysis that force units, periods, definitions and shown calculations, plus a two-minute tie-out check.
Optimise a prompt
How to improve an AI prompt systematically: success criteria, test cases, one change at a time. Plus what prompt optimizer and enhancer tools do, and don't.
Prompt engineering guide
Prompt engineering guide: the techniques Anthropic, OpenAI and Google still recommend, which 2023 habits to drop, and examples to copy for analysis and research.
Prompts that cite sources
The grounding techniques Anthropic, OpenAI and Google all recommend, with copy-paste prompts and a two-minute check for the citations you get back.
System prompts
What a system prompt is, where it lives in Claude, ChatGPT and Gemini, what Anthropic's guide says for the newest Claude models, and an example to copy.
AI accuracy
AI benchmarks explained
What GPQA, MMLU-Pro, SWE-bench, Humanity's Last Exam and the big leaderboards measure, why many are saturated or contaminated, and how to read a score.
AI hallucination rates
AI hallucination rates run from under 2% to over 80% depending on the task. The dated figures, what each one measures, and how to measure your own.
Best AI for data analysis
Which AI is most accurate for data analysis, what spreadsheet and SQL benchmarks show, and the checks that catch a wrong number before it reaches a dashboard.
Best AI for finance
How accurate AI is on real finance tasks, what FinanceBench and Vals AI measure, and a two-minute check before an AI number reaches a model or memo.
Best AI for research
Independent studies of AI search and deep research tools, what their citation accuracy really is, and how to test a research AI on a topic you know.
How to check AI answers
How to fact-check ChatGPT, Claude or Gemini: split the answer into claims, sort them by type, and run the right check on each. With a worked example.
Most accurate AI
Which AI is most accurate depends on what you measure. Three leaderboards, three different winners, and a two-minute test to run on your own questions.
AI for your role
AI for CRM data
Connect Claude to HubSpot, Salesforce or Attio in four steps, then avoid the traps that make CRM answers wrong: duplicates, stale stages, currency and page limits.
AI for FP&A
Set up Claude for budgeting, forecasting, variance and scenario work: a project with your definitions, Claude for Excel, read-only actuals, and checked prompts.
AI for operations
Set up Claude for an operations team: the context it needs, connectors for Jira, Confluence, Notion and Slack, and five AI ops workflows with full prompts.
AI for process improvement
Use Claude to map a process from ticket and event logs, find where work waits, test competing root causes and measure before and after, with setup and prompts.
AI for product managers
Set up AI for product management: the context Claude needs, official connectors for Amplitude, Linear, Jira and Intercom, and five PM workflows with full prompts.
AI for RevOps
Set up AI for revenue operations: the context Claude needs, official connectors for Salesforce, HubSpot and Stripe, and five RevOps workflows with full prompts.
AI prompts for PMs
Copy-paste AI prompts for product managers: discovery, PRDs, acceptance criteria, metrics and experiments, each built to show its evidence, plus a check.
AI SOP generator
Set up Claude as an SOP generator: a project, a skill with your template, connectors to Granola and Confluence, a copy-paste prompt, and the walk-through check.
Best AI tools for finance
The 10 best AI tools to run alongside Claude in a finance team, ranked, with a winner, what each gets wrong, and which to pick for your team's shape.
Best AI tools for marketing
The 12 best AI tools to run alongside Claude in a marketing team, ranked, with a winner, what each gets wrong, and which to pick for your team's size.
Best AI tools for PMs
The 10 best AI tools to run alongside Claude as a product manager, ranked, with a winner, what each gets wrong, and which to pick for your team's shape.
ChatGPT for finance
How to use ChatGPT for finance and accounting in 2026: ChatGPT for Excel, the QuickBooks, Xero and Stripe apps, three workflows, and the checks before figures ship.
ChatGPT for marketing
How to use ChatGPT for marketing in 2026: the features to use, three full workflows with prompts and example outputs, and the checks before anything ships.
Claude for finance
Set up Claude for a finance team: Claude for Excel, connectors for Xero, QuickBooks and Stripe, and five AI finance and accounting workflows with full prompts.
Claude for marketing
Set up Claude for a marketing team: the context it needs, official connectors for GA4, Google Ads and HubSpot, and five AI marketing workflows with full prompts.
Claude skills for marketing
A complete Claude skill for a marketing report, how to install it step by step, what Anthropic's marketing plugin includes, and how to test a skill first.
AI for data and analytics
AI business intelligence
What the AI in Power BI, Tableau, Looker, ThoughtSpot, Omni, Sigma, Hex, Metabase and Count does, which to choose, and how to check its answers.
AI dashboards
What AI dashboard generators in Claude, Power BI, Looker and Omni really do, how to build one where every number traces to a query, and a two-minute check.
AI data analyst
What an AI data analyst does well, where realistic benchmarks show it failing, whether it can replace an analyst, and how to review its work like a new hire's.
AI for data analytics
Set up Claude for a data team: the context it needs, official connectors for Snowflake, BigQuery and dbt, and five AI data analytics workflows with full prompts.
AI report generator
How AI report generators in Claude, ChatGPT and Power BI work, a project setup where every number is footnoted to a query, and the sentences to check first.
Best AI tools for data teams
The 9 best AI tools to run alongside Claude on a data team, ranked, with a winner, what each gets wrong, and the setup for your team's shape. Checked October 2026.
BigQuery MCP
Google's BigQuery MCP server: its six tools, how to connect it to Claude, how to block writes with IAM, the 3,000-row cap, and how to check the answers.
Click-through rate (CTR)
CTR = clicks ÷ impressions. Google Ads, Search Console, Meta and email each count both differently, which is why your CTRs don't match. Worked examples and checks.
Conversational analytics
What conversational analytics is, the two ways chat-with-your-data tools build answers, where they guess at definitions, and a two-minute check for each answer.
Conversion rate
Conversion rate = conversions ÷ visits. Google Ads, GA4 and your store each count visits differently, which is why their rates don't match. Worked example and checks.
Cost per acquisition (CPA)
CPA = ad spend ÷ conversions. Which conversions count, and which platform claims them, is why Google Ads, Meta and your CRM disagree. Worked example included.
Customer acquisition cost
CAC = sales and marketing cost ÷ new customers. Which costs and customers you count can halve or double it. Blended and paid CAC, payback and a worked example.
Databricks MCP
Databricks' managed MCP servers (Genie, DBSQL and more): how to connect them to Claude, the Unity Catalog grants to set, the risks, and how to check answers.
dbt MCP
dbt's official MCP server, remote and local: its tools, how to connect it to Claude, which defaults to change before you do, and how to check the answers.
Postgres MCP
Postgres has no official MCP server and the reference one has a SQL-injection hole. What to use instead, how to connect it to Claude, and how to lock it down.
Return on ad spend (ROAS)
ROAS = conversion value ÷ ad spend. The formula is simple; the four choices inside it are why your ad platform and your finance team report different numbers. Break-even table included.
Semantic layer for AI
Why AI answers about your data need a semantic layer, what goes in one, how dbt, Snowflake, Looker and Power BI compare, and how to connect yours to Claude.
Snowflake MCP
How Snowflake's managed MCP server works, how to connect it to Claude step by step, which roles and grants to set, and how to check what it returns.
Text-to-SQL
How text-to-SQL works, what Spider 2.0 and BIRD really show in 2026, why semantic layers matter, and prompts and a two-minute check for AI-written SQL.