AIAP Topic Courses·Using AI for Research·35:53

Evaluating AI Research Tools, Part 2: Buyer Personas, Marketing Trends, a Content-Gap Analysis, the Verdict, and Adapting a Proven Prompt

Outskill instructor (unnamed) Research trainer - runs identical prompts through six AI research tools and scores them live

▶LMS lesson

The short version

Use case 3 - four Notion buyer personas (demographics, psychographics, jobs to be done, pains, buying behaviour, channels, objections, voice of customer): runs of 30-35 minutes; ChatGPT 'so far the winner'; Gemini stopped on an error once but has the best paragraph-level sources; a '37.14% aged 25-34' stat checked out (0:00-0:14).
Use case 4 - emerging trends in AI-led marketing, Perplexity added: Perplexity fastest and 4/4/4-4.5; ChatGPT and Gemini slowest; Claude deviated from the prompt; Genspark too brief. Use case 5 - content-gap analysis for Fireflies.ai vs Otter, Rev, Trint, Grain, AssemblyAI: Manus surprised with an exportable 90-day calendar; Gemini best on themes and search; Claude missed Granola (0:12-0:34).
Verdict: Gemini Deep Research first, ChatGPT second, Claude and Perplexity decent, Genspark best presentation least depth; research in Gemini, augment in ChatGPT, format in Genspark. Workflow: paste a proven prompt and ask the LLM to adapt it to a new subject (Vapi) - 80% of a perfect prompt, then refine (0:31-0:35).

At a glance, three clicks deep

Skim here first: the closed row is the glance, open is the study card with the key points and timestamps, and the ↓ link drops to that concept's full write-up below.

01Use cases 3-5: personas, trends, content gapsScores vary by task;›

Scores vary by task; always verify a statistic.

Personas (0:00-0:14)

Trends (0:12-0:23)

Content gap (0:23-0:34)

↓ Full write-up of this concept

02The verdict, the recommended stack, and adapting a proven promptGemini -> ChatGPT -> Genspark;›

Gemini -> ChatGPT -> Genspark; adapt prompts, don't rewrite.

Verdict (0:31-0:32)

Adaptation (0:33-0:35)

↓ Full write-up of this concept

The concepts in full

Use cases 3-5: personas, trends, content gaps

Manus, the messy one in round one, handed over a 90-day content calendar as a spreadsheet in round five. Rankings move with the task.

Personas (Notion): ChatGPT ~4/4.5/4-4.5; Gemini 4/4/4 with praised source UX but one stalled run; Claude 4/3.5/4; Manus 3.5/4.5/4 with only ~10 references; Genspark 3.5/4/4.5; a 37.14% age-band statistic verified true. Trends in AI-led marketing (categories, frameworks, adoption timeline, enablers, sector impact, M&A, geography), Perplexity added: Perplexity fastest, 4/4/4-4.5; ChatGPT and Gemini slowest; Claude strayed from the brief (action 3); Genspark 3.5 across. Content gap for Fireflies.ai vs Otter, Rev.com, Trint, Grain, AssemblyAI over the last 12 months (strategy, gaps, positioning, formats, SEO keywords, prioritisation, calendar): Claude top-5 recommendations but missed Granola; ChatGPT strongest on brand voice and cadence; Gemini best on themes and organic search; Manus most actionable with an exportable 90-day calendar; Genspark nine pillars only; Perplexity a good graph, shallow plan.

Why it matters

Three more business prompts and the pattern that no tool wins every time.

The verdict, the recommended stack, and adapting a proven prompt

Research in Gemini, augment in ChatGPT, format in Genspark. And never write the second prompt from scratch - hand the LLM the first one.

Verdict (mid-2025, subjective): Gemini Deep Research first choice, ChatGPT Deep Research second, Claude and Perplexity decent, Genspark best presentation with least depth. Stack: research in Gemini, augment with ChatGPT, format in Genspark. Prompt adaptation: paste a proven prompt and say 'I want to do this research for Vapi - modify the prompt to work for Vapi'; the LLM swapped industry (voice AI agents) and competitors (ElevenLabs, AssemblyAI, Deepgram, OpenAI Whisper, Speechify). Let the LLM draft the baseline prompt to reach '80% perfect', then refine with domain expertise. The course promises all five prompts in the resources.

Why it matters

Turns one good prompt into a reusable research instrument.

This idea elsewhereclosesThe method

Tools referenced

Tool	Coverage	Context
Gemini	demonstrated	Deep Research - first choice
ChatGPT	demonstrated	Deep Research
Perplexity	demonstrated
Manus	demonstrated
Genspark	demonstrated
Claude	demonstrated	Claude Research
Fireflies	mentioned	Content-gap subject

Action items

Resources mentioned

Resources

docThe five research prompts (market landscape, Notion benchmark, Notion personas, AI-led marketing trends, Fireflies content gap)

Extraction notes

This page was built from an auto-generated transcript, which garbles product and people's names. Those were corrected silently in everything above and logged here for transparency. The warnings flag claims that were true on the recording day but change fast.

Transcript corrections applied

The transcript says	The trainer actually means
Print Grain / Simpli AI	Trint, Grain / AssemblyAI
WAPI / PAPI	Vapi
publicity	Perplexity
cloud / Cloud co work / cloud research	Claude / Claude Cowork / Claude Research
chat g b t / ChargeGPT / Chargebee / chat GBT	ChatGPT
ATN / n a 10 / Anytime	n8n

True on recording day — verify before relying

Mid-2025 tool rankings and speeds (Gemini Deep Research first, ChatGPT second, Genspark best presentation) are subjective and have likely shifted; the method is evergreen