Tip to try out

If you want to automate your work on LinkedIn, you should take a close look at three providers in particular: Expandi.io, Waalaxy and PhantomBuster. Each of the tools has its strengths and weaknesses and covers a specific task area very well. All three can be automated via APIs and integrated into end-to-end processes. At their core, they are used to automatically approach LinkedIn members for marketing and sales purposes. Used wisely, they make sales work easier without getting into legal hot water and inconveniencing LinkedIn members.

LinkedIn as a global address book

The social network LinkedIn, which belongs to the Microsoft Group, is the world's leading address book for salespeople, marketing experts and headhunters. On no other platform can you find so much professional information about a person. It is therefore not surprising that LinkedIn is one of the most popular contact sources for outbound activities alongside address databases such as ZoomInfo, Dealfront or Lusha. Since LinkedIn, like any social network, earns its money from advertising on its platform, it makes it difficult for its users to use the data stored in it externally. The aim is to retain its (preferably paying) members on the platform.

Making contact to start the dialog

Consequently, all communication with each other should also take place on LinkedIn. This is only possible without restriction (especially for non-paying members) if you have "connected" with the person you are talking to. The means of choice is the contact request, which is inherent in every profile.

If the addressee of the contact request has accepted it, messages can be exchanged as desired. Otherwise this option is not available. Paying members also have the option of exchanging a certain number of InMail messages with non-contacts. However, the number is limited each month. LinkedIn favors knowing who knows whom. This allows more specific advertising to be placed in the best possible way.

In a nutshell: Making contact is the gateway to dialog with other members.

Create personalized LinkedIn contact requests with AI

To maintain its attractiveness, LinkedIn limits the number of contact requests. This is to prevent spam. This limit varies based on the company's own Social Selling Index, which is intended to reflect its own attractiveness as a LinkedIn member. The actual value obviously varies between 20-25 invitations per day - or less.

Standard or personal 

Consequently, successful salespeople must first establish and expand their contact base in order to present their concerns in the second step.

LinkedIn offers two types of contact request.

• A standard text without a personal touch

• A personal message with a maximum length of 300 characters

The chance of a stranger accepting a contact request is of course much higher with personalized contact requests. Especially if your LinkedIn profile "smells" like sales or marketing, you should choose this route.

However: What text should you write in the invitation?

Best practice for invitations

The invitation text should be primarily relevant. The invitee should find themselves in it, as they legitimately expect a thematic connection or even a common interest with the inviting party. Or do you want to be able to connect with everyone on social networks?

The most effective source for a relevant invitation text are the posts, comments or published articles of the invitee. What are they interested in? What do they have in common? What demand might they have that I can meet?

This requires human intelligence and human decision-making - and time. This process is time-consuming.

Generative AI automates invitation texts

Thanks to Generative AI, a lot of work can be automated here. Nothing is faster at understanding and synthesizing longer texts than large language models. They penetrate the meaning and content of posts, comments and entire articles in a matter of seconds. They summarize them succinctly. They are able to write specific invitation texts. Bingo!

With the help of OpenAI's GPTs (specialized ChatGPT bots) or agents (GPTs that can be addressed via API), this task can be completely delegated to artificial intelligence. The AI model is given the input of the person to be contacted and receives a relevant, well-formulated contact request that can be sent immediately to the addressee.

For this to work accurately and error-free, a targeted parameterization of the AI model - in this case the GPT or agent - is necessary. This is the secret to success. Enriched with specific information about the company and its main areas of activity (for technicians: retrieval augmented generation plays a role here), the AI model selects posts with similarities or high relevance in order to formulate a specific invitation text.

Note: Generative AI really shows its strengths in text analysis and text creation.

Successful LinkedIn automation in the sales process

That takes care of the most important point: the automated formulation of personalized contact requests. How can this be scaled? We ultimately want to add a whole list of interesting people to our LinkedIn network!

Process automation works wonders

This is where process automation comes into play. The recipe is as follows:

1. Create a list of interesting contacts using tools such as LinkedIn Sales Navigator, Dealfront, Apollo.io or other contact databases.

2. Then use the tools described above to add the contributions of interesting contacts to this list.

3. Now let OpenAI's GPT model create a relevant personalized invitation text based on the collected contributions and the name of the contact.

4. One of the LinkedIn automation tools mentioned above may send this invitation text to the list of interesting contacts mentioned under 1.

5. If a person accepts the invitation, further automated messages can be sent - or the human takes the wheel.

Workato, Make and Zapier

Any powerful integration and automation platform can be used to automate this process: Workato, Make, Zapier - or even self-created Python programs or programs created with ChatGPT. It orchestrates the process, ensures the correct mappings and calls as well as the step-by-step further processing of the information.

Of course, this can also be linked to a CRM system in order to initiate follow-up steps and operate structured account and lead management.

The entire process can be expanded and adapted as required. Sales, marketing and headhunting activities no longer have to be laboriously performed manually if they can be automated to the same quality. Of course, as always, it depends on the detail and experience with AI-based automation processes. However, the AI platforms currently available on the market can be used in conjunction with process automation for AI-supported contact tracing in LinkedIn.

Artificial intelligence helps people to increase their productivity - even in personalized sales.