30 Practical Ideas for Business Development Managers to Stay Cognitively Sovereign
AI prospect research tools like Apollo.io and LinkedIn Sales Navigator can surface contact lists quickly, but they miss the relationship history and organisational politics that actually move deals forward. Without deliberate practise, your instinct for which opportunities are worth pursuing atrophies. These 30 ideas help you stay in control of your commercial judgement while using AI to handle the mechanical work.
These are suggestions. Take what fits, leave the rest.
Write a one-sentence relationship hypothesis before running an AI searchbeginner
Before Apollo.io surfaces a prospect list, write down what problem you think this organisation has and how you know someone there who might care. After AI returns results, check whether it found the person you predicted or someone different.
Use ChatGPT to explain why an AI-suggested prospect matters, then call someone who knows themintermediate
When AI flags a contact based on job title or company size, ask ChatGPT to explain the connection. Then call a peer or former colleague and ask whether that reasoning actually holds water in practice at that organisation.
Track which AI-researched prospects came from referral introductions versus cold listsbeginner
In your CRM notes, mark whether each contact was suggested by AI research or came through a warm introduction. After three months, compare conversion rates. This shows you whether AI-generated lists need relationship scaffolding to convert.
Spend 15 minutes on a prospect's social media before reading the AI summarybeginner
Open LinkedIn and scroll through a prospect's recent posts, comments, and shares before letting Perplexity or ChatGPT synthesise their profile. Notice what you spot that AI missed: their stated priorities, the people they engage with, the problems they care about.
Ask AI for three possible relationship entry points, then choose one based on your networkintermediate
When Sales Navigator suggests a contact, ask ChatGPT to propose three ways you might connect with them. Pick the entry point where you actually know someone, not the one AI ranks highest.
Build a shadow research file on one key account without using AI toolsintermediate
Choose one strategic account and research it manually for one week using only LinkedIn, company news, and your own network calls. Note what you discover about decision-makers, recent changes, and relationship dynamics. Compare your findings to what AI would have surfaced.
Keep a list of prospects AI rejected or ranked low that you closed deals withbeginner
When you win a deal, check whether Apollo.io or LinkedIn Sales Navigator had flagged that prospect. Document the ones AI missed or deprioritised. After six months, you will see patterns in what AI overlooks.
Talk to your sales team about prospects AI kept resurfacing that never movedbeginner
In your weekly sync, ask colleagues which AI-suggested prospects they kept pursuing but never engaged. Those patterns tell you what criteria Apollo.io or Sales Navigator is using and where those criteria fail.
Write a research brief for AI using a prospect contact you landed through a referralintermediate
After closing a warm introduction to a key contact, feed that person's profile to ChatGPT or Perplexity and see how thoroughly it understands them. Note the gaps between the AI summary and the actual conversation you had.
Before uploading a list to Salesforce Einstein, review it manually and predict which names will convertundefined
When AI generates a prospect list, print it or open it in a spreadsheet. Spend 20 minutes going through each name and marking which ones you think will actually open an email from you. Later, compare your predictions to actual open rates.
Outreach: Writing Emails That Reflect Your Actual Thinking
Write your subject line first, without AI, then let ChatGPT suggest three alternativesbeginner
Before using AI to draft an outreach email, write the subject line yourself. It should reference something specific about that prospect's business. Then ask ChatGPT for three options and compare them to your original.
Track open and reply rates on emails you wrote versus AI-generated sequencesbeginner
Tag your outreach emails in your CRM to mark which were written by you and which came from Apollo.io or ChatGPT sequences. After 20 sends of each, compare open rates and reply rates. This shows you where AI sounds like everyone else.
Every third email in your sequence should reference a specific conversation you hadintermediate
When building a multi-touch campaign, let AI draft emails two and four, but write emails one, three, and five yourself. In your email, mention something you learned on a call with a peer or customer in that prospect's industry.
Use AI to rewrite your email for tone, not for strategybeginner
Draft your outreach message first with your own thinking about why the prospect should care. Then paste it into ChatGPT and ask it to tighten the grammar and tone. Do not ask it to reframe your core message.
Save the best email you ever wrote and feed it to AI as a style referenceintermediate
Find an email that got a reply from a prospect you respected. Copy that email into ChatGPT and tell it to use this style as a reference for future outreach. This anchors AI output to your actual voice.
Spend one week sending only manually written emails, then compare results to AI weeksintermediate
In one week, write every outreach email yourself without using ChatGPT or Apollo.io sequences. In the following week, use AI drafts. Compare reply rates, meeting request acceptance rates, and conversation quality.
Ask prospects directly whether your email felt personal or genericintermediate
On calls with prospects who replied to your outreach, ask them what made them open that email. You will hear whether they saw something specific to them or whether it felt like a mass campaign.
When AI-generated emails underperform, rewrite one sentence that does the real workintermediate
If an AI-drafted email gets no replies, look at what you know about that prospect that AI did not write about. Rewrite one specific sentence that mentions something personal. Test that version with similar prospects.
Build an email template from your recent wins, not from AI suggestionsintermediate
Open the emails you sent to prospects who became customers. Extract the common elements: how you opened, what problem you mentioned, how you asked for the next step. Build a template from that pattern, then use AI only to vary the specifics.
Write the opening line of every email before pasting it into an AI toolundefined
The opening line should reference something specific about that prospect: a recent hire, a company announcement, a shared connection. Only after you write this first line should you use ChatGPT to draft the rest.
Opportunity Assessment: Keeping Your Instinct Sharp
Before running a deal through Salesforce Einstein, write your own stage prediction and reasoningbeginner
When a new opportunity comes in, write a quick note about where you think it stands and why. Include your gut read on whether this deal is real or just a conversation. Only then ask AI to assess it.
Keep a decision log of opportunities you moved forward despite AI flagging them as low probabilityintermediate
When an opportunity has low scores from Einstein or ChatGPT analysis but you pursue it anyway, note the reason. Track how many of these hunches become real deals. This trains your instinct against the AI baseline.
Call your best deal-maker colleague and get their read before trusting an AI opportunity assessmentbeginner
When a deal looks marginal based on AI analysis, call someone in your organisation who has a strong track record closing deals. Ask them whether the prospect profile and signals match deals they have actually won.
Track which deals AI said would not close but your relationship management turned into winsintermediate
In your CRM, tag deals that had weak AI probability scores but closed because of relationship effort. After a quarter, count how many wins came from relationships that looked unpromising on paper.
Ask prospects why they decided to move forward, then compare their reason to the AI assessmentintermediate
After a prospect agrees to proceed, ask them what moved the needle. Write their answer in your CRM. Compare it to what Salesforce Einstein or ChatGPT said would matter. You will see what AI gets wrong about real motivation.
Build a personal scorecard for opportunity assessment separate from AI rankingsintermediate
Create your own simple scorecard: relationship strength with the main contact, clarity of their problem, timing pressure, budget ownership, competitive alternatives. Score opportunities on your card alongside Einstein scores. Over time, see which correlates with wins.
When an AI assessment conflicts with your read, schedule a conversation instead of accepting the AI scorebeginner
If Salesforce Einstein says a deal is early stage but you sense urgency from the prospect, call them. Let that conversation reset your assessment, not the algorithm's score.
Review opportunities that stalled and see whether AI correctly predicted whyintermediate
When a prospect stops responding, look back at your notes and the AI assessment. Did Einstein correctly identify the risk factors? Did you miss signals that AI caught? This tells you whether you can trust AI reads on similar deals.
Spend 30 minutes per month assessing one opportunity with zero AI inputbeginner
Choose one live deal and analyse it purely from your own knowledge, conversations, and instinct. Write your stage prediction, risk assessment, and next steps without consulting any AI tool. Later, compare your assessment to what AI suggested.
Keep a wins analysis document that lists what AI said versus what actually matteredundefined
After each closed deal, write one line about the AI assessment (what Salesforce Einstein flagged, what ChatGPT analysis suggested) and one line about what actually moved the deal (relationship moment, budget trigger, competitive pressure). Over time, you build your own model of what works.
Five things worth remembering
The deals you close through relationships you built yourself will always tell you more than the deals AI algorithm says will close. Pay attention to that signal.
When AI-generated outreach feels impersonal to you as you read it, it will feel impersonal to the prospect too. Trust that reaction.
Your best days as a business development manager came from instincts you built over years of deal-making. AI tools should expand your capacity, not replace your reasoning.
Every month, spend one full day working the way you did before you had these AI tools. It keeps your commercial instinct from atrophying.
The prospects who reply to you are telling you something about what works. Listen to them more than you listen to what AI tools say should work.