Case Study: Sales Team Uses AI Research Agents to Close 35% More Deals

Company: OnboardHQ (pseudonym), B2B employee onboarding SaaS Team size: 5-person sales team (4 AEs, 1 SDR) Challenge: Reps spending 3+ hours/day on prospect research instead of selling Result: Research time cut to 20 minutes/day, close rate up from 18% to 24%, pipeline velocity increased 35%

---

Sales reps know they should research prospects before every call. But in practice, researching a single prospect thoroughly -- company background, recent news, key decision-makers, likely pain points, competitive landscape -- takes 30–45 minutes. For a rep making 8–10 calls per day, that's 4–6 hours of research time. Time not spent selling.

OnboardHQ's sales team was caught in this trap. Their reps averaged 3.2 hours per day on prospect research, leaving only 4.8 hours for actual selling activities (calls, demos, follow-ups). The research was necessary -- reps who researched thoroughly closed at 22% versus 14% for unresearched prospects -- but the time cost was unsustainable.

They automated prospect research with an AI agent squad on Ivern. Now, a comprehensive prospect brief takes 90 seconds to generate. The team's close rate improved from 18% to 24%, and reps gained 2+ hours per day for selling.

Related: AI for Sales Teams: Complete Guide · How to Build an AI Sales Outreach Squad · AI Agent Sales Outreach Workflow · How to Automate Research with AI Agents

The Research Bottleneck

OnboardHQ sells to HR leaders at companies with 200–2,000 employees. Each prospect requires research across multiple dimensions:

Research Area	Time	Purpose
Company overview (revenue, size, industry)	10 min	Understand their context
Recent news (funding, layoffs, growth)	10 min	Identify timely talking points
Tech stack analysis	5 min	Integration opportunities
HR technology landscape	10 min	Understand current tools
Key decision-makers	10 min	Know who to contact and how
Competitor usage	5 min	Differentiation angles
Total per prospect	~50 min

With 4–6 prospects per day, research consumed 3–4 hours. Reps were choosing between thorough research and more selling time -- a lose-lose trade-off.

The AI Prospect Research Squad

OnboardHQ's sales ops lead built a 3-agent squad that generates comprehensive prospect briefs from a company name and website URL.

Agent 1: Company Intelligence Researcher

• Model: Gemini 2.5 Pro (free tier)
• Role: Deep-dive company research
• Prompt:

  "Research [company name] ([website URL]). Provide: (1) Company overview: founding year, revenue estimate, employee count, industry, business model. (2) Recent developments: last 6 months of news, funding rounds, product launches, executive changes, layoffs, or growth announcements. (3) Technology stack: identify tools they likely use based on job postings, website, and public information. (4) HR technology: what onboarding, HRIS, and people tools they currently use or have used. (5) Key decision-makers: identify the likely VP of HR, Head of People, or CHRO based on LinkedIn/public data."

Agent 2: Pain Point Analyzer

• Model: Claude Sonnet 4
• Role: Identify likely pain points and map to product value
• Prompt:

  "Based on the company research, identify: (1) Likely pain points related to employee onboarding, given their size, growth rate, and industry. (2) Specific challenges they may face: compliance requirements, remote onboarding complexity, high-volume hiring, international employees. (3) How [product name] addresses each pain point -- provide specific feature-to-problem mappings. (4) Recommended conversation angles and opening talking points. (5) Potential objections based on their current tech stack and likely priorities."

Agent 3: Briefing Writer

• Model: Claude Haiku
• Role: Compile research into a scannable one-page prospect brief
• Prompt:

  "Synthesize the company intelligence and pain point analysis into a one-page prospect brief formatted as: COMPANY SNAPSHOT (3 lines), KEY TALKING POINTS (3–5 bullets), LIKELY PAIN POINTS (ranked by probability), PRODUCT FIT (top 3 features that map to their needs), POTENTIAL OBJECTIONS (with suggested responses), RECOMMENDED APPROACH (1–2 sentences). Keep under 400 words. Designed to be read in 60 seconds before a call."

The Workflow

SDR identifies prospect
    ↓
Enters company name + URL into Ivern task
    ↓
Company Intelligence Researcher (45 seconds)
    ↓
Pain Point Analyzer (30 seconds)
    ↓
Briefing Writer (15 seconds)
    ↓
One-page prospect brief delivered (90 seconds total)
    ↓
Rep reviews brief (60 seconds) and makes the call

The SDR runs 6–8 prospect briefs each morning in about 15 minutes of setup time. The agents generate all 6–8 briefs in parallel while the SDR does other work.

Results After 4 Months

Time Savings

Metric	Before	After	Change
Research time per prospect	50 minutes	2.5 minutes	-95%
Daily research time (team total)	16 hours	1.3 hours	-92%
Daily selling time (team total)	24 hours	32 hours	+33%
Prospects researched per week	40	80	+100%

Sales Performance

Metric	Before	After	Change
Close rate (researched prospects)	22%	28%	+27%
Close rate (all prospects)	18%	24%	+33%
Average deal size	$14,200	$15,800	+11%
Pipeline velocity (days to close)	45	35	-22%
Monthly revenue	$186,000	$251,000	+35%

The close rate improvement came from two sources: more consistent research (every prospect gets researched now, not just the ones reps have time for) and higher-quality research (the AI consistently covers areas human researchers miss).

Prospect Research Quality

The sales manager compared AI-generated briefs to manually researched briefs:

Aspect	Human Research	AI Research
Company overview accuracy	85%	92%
Recent news coverage	60%	88%
Pain point relevance	75%	70%
Decision-maker identification	50%	65%
Preparation time	50 minutes	2.5 minutes
Consistency across reps	Low	High

Cost

Item	Monthly Cost
Gemini 2.5 Pro (research)	$0.00 (free tier)
Claude Sonnet 4 (pain point analysis)	$8.00
Claude Haiku (brief writing)	$1.20
Total monthly cost	$9.20
Revenue impact	+$65,000/month
ROI	7,065:1

Why AI Research Outperformed Manual Research

1. Consistency

Every prospect gets the same depth of research. Before, research quality varied by rep, by time of day, and by how many prospects were in the queue. The AI doesn't get tired or cut corners.

2. Coverage

The AI consistently checks areas that humans skip when rushed: competitor usage, tech stack implications, and compliance requirements relevant to the prospect's industry. This broader coverage surfaces more relevant talking points.

3. Speed Enables More Research

Because a brief takes 2.5 minutes instead of 50, the SDR now researches every single prospect -- not just the "high-value" ones. This led to unexpected wins: several mid-value prospects that wouldn't have been researched manually turned into closed deals.

4. The Pain Point Analyzer Connects Research to Selling

Raw research is just data. The Pain Point Analyzer agent transforms data into selling insights -- specific feature-to-problem mappings, conversation angles, and objection responses. Reps don't just know about the prospect; they know how to talk to them.

Challenges

1. Rep Adoption Required Proof

Sales reps are naturally skeptical of tools that claim to save time. The sales manager ran a 2-week trial: one team used AI briefs, the other continued manual research. The AI team's close rate was 5 points higher. After that, adoption was immediate.

2. Brief Accuracy Varies by Company Size

For companies with limited public information (small private companies), the AI sometimes makes educated guesses. Reps learned to verify key claims during discovery calls rather than assuming everything in the brief is confirmed.

3. Personalization Still Matters

The AI brief is a starting point, not a script. The best-performing reps add their own relationship context -- "I noticed you went to [same university]" or "We worked with [similar company] last quarter" -- to the AI-generated talking points.

Build Your Sales Research Squad

1. Sign up free at ivern.ai/signup
2. Add API keys -- Google (free tier for research) and Anthropic ($5 credit)
3. Create a 3-agent prospect research squad with the roles above
4. Customize pain point prompts for your specific product and market
5. Run your first 10 prospect briefs and compare to your manual research

Ready to close more deals? Create your sales research squad →

---

This case study is based on aggregated patterns from B2B sales teams using Ivern AI for prospect research automation. Results represent typical outcomes for teams of 3–8 AEs. Individual results vary based on market, product complexity, and sales process maturity.