Cold outreach rule I wish I'd learned earlier: ask a question they can answer in one line.
Big asks get ignored. Small ones get replies.
What 8 makers on MakerMap learned about it, in their own words: their best posts on X, filed under Marketing.
Cold outreach rule I wish I'd learned earlier: ask a question they can answer in one line.
Big asks get ignored. Small ones get replies.
Spam test on my new email domain: 1.1/10 💀
So instead of sending 100 emails, I sent 5. Each one written for one person.
3 opened within 2 hours.
Slow is fast.
How to build outbound that actually books meetings:
1. Send to signals, not lists.
Funding rounds, hiring spikes, tech changes - those are the moments.
2. One well-timed message beats ten generic ones.
3. Write to the context, not the title.
Same role, different timing, completely different reply rate.
4. Automate the research.
Own the thinking.
5. Let the inbox tell you what's working.
Replies are feedback. Treat them that way.
6. Score before you sequence.
Prospects are ready at different times.
7. Improve the copy continuously.
The first version is a starting point.
Been building Hilead around these ideas - AI that detects the signal, finds the contact, writes the message, and books the meeting.
Most outbound fails before it's even sent. Wrong person, wrong moment, and the timing kills the message before it lands.
we got rid of fancy email templates at @graphify and now only send plain emails directly from our founder's account @safishamsii and he replies to every single email btw
this alone has boosted our delivery to open ratio significantly 🚀
Cold emails that get ignored all make the
same mistake they pitch too early.
→ Line 1: prove you looked at their business
→ Line 2: name their exact problem
→ Line 3: offer the fix in one sentence
→ Delete everything else
Short beats clever every time.
#coldemail #saas
We sent cold emails to property managers today and followed up on earlier outreach.
Pitching cold inboxes takes follow-ups before an operator opens the note.
Asking mutual contacts for introductions puts Property Copilot directly in the inbox.
Grok Bot + might be one of the most underrated combinations for finding customers right now.
Forget buying another list of 10,000 companies from Apollo and sending the same cold email to everyone.
You can now find people actively experiencing the exact problem your product solves. Not based on their job title or company size, but on what they're actually discussing, complaining about, or looking for.
Here are 10 prospecting plays worth trying:
1. Complaining about competitors → Perfect timing
Find people frustrated with a competitor's pricing, missing features, or poor support. You already know what problem to address.
2. Asking for recommendations → Buying intent
Find conversations like "What's a good alternative to X?" or "Can anyone recommend a tool for Y?"
3. Hiring for repetitive work → Automation
Find companies hiring for manual data entry, reporting, research, or administrative tasks your product could automate.
4. Recently funded → New budgets
Find startups that raised funding in the last 90 days, identify what they're hiring for, and match that against your offer.
5. Negative reviews → Switching opportunities
Find recent complaints on G2, Capterra, Reddit, or social media. Look for frustrations your product actually solves.
6. New leadership → Changing priorities
Find companies that recently appointed a CTO, CMO, or Head of Sales. New leadership often means new tools and processes.
7. Product launches → Missing infrastructure
Find startups launching on Product Hunt, Hacker News, or X. Identify gaps in analytics, onboarding, documentation, or other infrastructure.
8. Business expansion → New needs
Find companies entering new markets or opening locations. Growth often creates problems their existing tools can't handle.
9. Competitor users → Feature gaps
Find people discussing competing products and struggling with specific integrations, workflows, or limitations.
10. Technical questions → Unsolved problems
Find developers asking how to implement something your product already provides.
Here's where it gets interesting.
Grok Bot helps discover conversations and intent signals on X. lets you research across additional sources, enrich prospects, and build targeted discovery workflows. Combine them with AI agents to automate much of the research and qualification process.
Instead of spending weeks building a massive lead list, test five different signals with 20 prospects each. See which ones convert, discard the rest, and scale what works. Then turn successful searches into recurring workflows that continuously discover new opportunities.
I'd rather have 50 prospects with a verified problem than 5,000 contacts who happen to match my ideal customer profile.
You no longer need a massive sales team to experiment with this. You just need to understand your customers well enough to know what signals to look for.
What prospecting play would you run?
working through our b2b sales journey right now: a prospect initially rejected our service because he wanted a full-time employee instead of an outside vendor.
we audited his company's website anyway to show what was missing, and after sending it over, he traded cold silence for his wechat contact.
still no final yes yet, but trading a rejection for a direct line means when the timing is right, we're the first call they make.
not a success story yet, but worth acknowledging the progress.