How To Leverage Machines for Awesome Outreach – Gareth Simpson

09 April 2019

Posted in: BSEO

Seeker’s very own MD Gareth Simpson offered this presentation on the power of machine learning and outreach. He shared sharing how the Seeker team scales outreach activities, allowing us to have more conversations with the right people, and ultimately win more links.

Overview

What was the talk about?

Gareth talked us through how to use machine learning (ML) to improve the outreach lifecycle. By using the bots to do the heavy lifting (without sacrificing the human elements of outreach) it’s simple to increase efficiency, have more conversations with the right people and win more links.

First, Gareth explained how ML fits into the complex web of artificial intelligence (AI). Ultimately, it can:

  • Classify data
  • Perform tasks without instruction
  • Improve its own performance
  • Automate and scale repetitive tasks
  • Emulate human decision
  • Classify content and production

Now that’s savvy robotics.

LMFAO: Leveraging Machines for Awesome Outreach

Then, we took a deep dive into the outreach process and how ML can help us reach new heights. Features include:

  • Keyword research
  • Ranking projections
  • Intelligent prospecting
  • Auto-routing
  • Pitch optimisation
  • Conversations

Fave quote

“Machine learning can be used to create more meaningful work and help you scale outreach campaigns without compromising human relationships.”

Potential impact on the industry

Outreach is hard – it’s no wonder both Kat Kynes and Shannon McGuirk discussed campaign fails at this April’s conference. It’s clear that to win links we need to get niche, accurate, and personal. But is that really scalable? With ML it is.

Gareth explains that ML can assist with:

Contact filtering & enrichment: Identify the person you want to get in front of, collect contact information (e.g. email address, social media handles), and mitigate human error (e.g. typos).

Processing data in inboxes: Read outreach emails, understand the urgency, sentiment, opinion and classify the semantics. Make a decision/action based on the text in emails.

Prospecting & list building: Scrape links, read page content, and analyse on-page attributes based on quality criteria to filter out unsuitable prospects. Classify the pages ready for manual inspection by a human team.

Media monitoring & email text extraction: Monitor PR tools, e.g. HARO, #journorequests, TalkWalker, keyword classify data, output grouped data for display (e.g. in a spreadsheet), and auto-respond.

Conversational: Employ psychology techniques in AI, auto-complete sentences, and suggest responses. (E.g. Gmail Smart Compose.)

And it will always feed back to the ML model when incorrect to make the process more accurate, efficient and intelligent.

LMFAO: Leveraging Machines for Awesome Outreach

Key takeaways

  • Identify the repetitive tasks: if it can be repeated, let ML do the heavy lifting to free you up for the human-only outreach tasks
  • Anyone can implement ML (not just devs): There are plenty of platforms already built, from full ML platforms and send optimisation to data prediction and text analytics. You just need to hook it up to you tools via the API. Simple.
  • Create more meaningful work: If you can let the bots identify the right sites, find contact details, select the appropriate subject line and triage your inboxes, you have more time to focus on building even more relationships and winning big links.
Share This