Most AI account research ends with a company summary. You get a few facts and generic talking points—then still have to work out what they mean for the decision in front of you.

Strawberry is a browser with built-in AI agents. Your companion can investigate real sites and logged-in tools, then combine what it finds with your CRM, emails, meetings, files, tabs, and connected apps. It helps you understand the account and decide what to do next.

Start with the question you need answered

Useful account research changes depending on the job. You might need to understand a company before reaching out, find a credible path into an account, prepare a strategy, learn which vendors or tools it uses, map decision-makers, or investigate one question before a meeting.

Tell your companion what you are deciding and how much time you have. Start with the most important question and go deeper when harder-to-find evidence could change the next step.

Bring your research and relationship context together

Account context rarely lives in one place. The CRM may hold the opportunity, email the relationship, and meetings the customer's own words. Useful company information may be spread across LinkedIn, company pages, job listings, documentation, databases, and industry-specific sites.

Your companion brings those pieces together without a giant prompt. It can also check whether the company is already a customer, past opportunity, or team connection before treating it as new.

Research the web more like a person would

Many AI research tools summarize the pages they retrieve. Because Strawberry works in the browser, it can open the source, follow links, search within sites, move across tabs, and continue when the first page falls short.

That matters when useful evidence lives outside the obvious company profile. Product and technical documentation, privacy or security pages, job listings, customer stories, niche directories, and public databases can reveal technology choices, initiatives, relationships, or changes that a standard data provider misses.

The goal is to know where else to look when obvious sources fall short—and when to stop.

Turn facts into a useful point of view

A strong brief separates facts from inferences, keeps links close to material findings, and treats likely priorities as hypotheses.

Depending on your question, the result might include:

  • What the company does and what has changed.
  • The relationship and relevant previous activity.
  • The people who matter and credible paths into the account.
  • Technology, vendors, initiatives, or market context.
  • Likely priorities or risks, clearly marked as hypotheses.
  • The next question or action that would move the work forward.

A few findings that change what you do beat a generic company profile.

Use feedback to make every account brief better

Tell your companion what mattered, what felt generic, which sources you trust, and what you want next time. It can carry forward your preferred depth, useful fields, qualification bar, and format.

When the brief would help others working the account, share an approved version while keeping private relationship context and unreviewed hypotheses separate. Share the accepted research method as a team skill when others should reuse it, or share the full companion when the team needs the same account context across several sales workflows. A Routine can watch agreed sources and prepare an update when something meaningful changes.

Connect research to the next piece of work

Use account research to prepare for a meeting, enrich missing contacts, send thoughtful outreach, update the CRM, or revisit the strategy. Strawberry carries the context into the next step.

Open the Research an Account skill with one company and the question you need answered. Your companion can start with the context already available, give you a useful first view, and go deeper where it will actually help.

Official Strawberry skill

Research an Account

Help the user understand an account well enough to make a better decision or take a useful next step. Adapt the depth and output to their question rather than producing the same company profile every time.

Workflow

1. Understand the question

Clarify which company, organization, or person the user means and what decision the research should support. They may need an overview, way into the account, current priorities, technology or vendor context, decision-makers, relationship history, or deeper work before outreach or strategy.

Agree on depth when time, cost, or breadth matters. Answer the most useful question first and go deeper when it could change the decision.

2. Start with what the user already knows

Use relevant CRM records, email, meetings, files, notes, tabs, memory, and connected apps before asking the user to repeat available context. Check whether the account is already a customer, prospect, partner, past opportunity, or known relationship. Existing history may matter more than a fresh public signal.

If the user only needs a concise brief for a scheduled meeting, use strawberry/sales/prepare-for-sales-meetings. Continue here when the account needs deeper investigation.

3. Investigate the sources that fit the question

Research the public web and relevant logged-in sources with a curious, evidence-led approach. Useful places include company and product pages, LinkedIn, news, job listings, customer stories, documentation, privacy and security pages, public databases, directories, and market-specific sources.

Because Strawberry works in the browser, follow links, search inside sites, compare pages, and keep digging when the first result is incomplete. Stop when the question is answered or more research is unlikely to change the decision.

4. Resolve identities and keep the evidence trustworthy

Confirm the company, domain, and people before combining records. Resolve ambiguity with multiple signals such as current role, employer, location, prior work, or an official company source.

Keep source links for material findings and dates where freshness matters. Separate facts, relationship context, and useful hypotheses. Surface meaningful contradictions or gaps.

5. Turn the research into a useful point of view

Prioritize what could change the user's next move. Depending on the goal, the result may include:

  • a concise account overview and what has changed;
  • the existing relationship and relevant prior activity;
  • likely priorities, needs, or risks clearly marked as hypotheses;
  • useful stakeholders, decision-makers, or paths into the account;
  • technology, vendors, initiatives, or market context relevant to the user's question;
  • questions worth answering next; and
  • a recommended next step with the evidence behind it.

Do not manufacture an outreach angle, force a recent detail, or turn every public fact into a sales signal. Prefer a few useful findings over a long generic profile.

6. Deliver and connect the next step

Return a concise, sourced brief in the format that best supports the decision: chat, document, table, CRM note, or another agreed destination. Lead with the quick take.

If the next step is a meeting brief, use strawberry/sales/prepare-for-sales-meetings. If the user needs missing contacts or fields, use strawberry/sales/enrich-lead-list. If they want to contact someone, use strawberry/sales/send-personalized-outreach. Keep research separate from external actions and ask before writing to a shared system.

When the approved brief would help others working the account, offer to share it with the relevant teammates. Keep private relationship context and unreviewed hypotheses out of the shared version.

After useful feedback, offer to save the accepted sources, depth, fields, output, and criteria as a custom or shared team skill. When teammates need ongoing access to the same account context and several workflows, offer to share the full companion. For an ongoing monitoring need, suggest a Routine that checks agreed sources for meaningful account changes and prepares a reviewable update.