Deberli Vision

Deberli helps a small B2B team get buying conversations from its own LinkedIn account and mailbox. It contacts the people with a reason to talk now first, and other people who fit your profile after them.

You tell Deberli who you sell to. Deberli finds the people who show a signal this week, and more people who fit your profile. It writes to them from your accounts and sends the follow ups. When a person wants to talk about buying, Deberli tells you. Each campaign shows which people accept, reply and buy, and the next campaign uses this data.

We measure Deberli with one number: buying conversations per person contacted.

The problem

In a small B2B team, nobody does outbound full time. The founder or a salesperson sends LinkedIn messages and emails between other tasks. Four things go wrong.

• They contact people at the wrong time. A person who fits the profile does not always need you this month. Most of these people do not reply.
• Messages and follow ups take too much time. The team writes each message by hand and forgets follow ups. When client work is busy, outreach stops.
• The tools count messages sent. Lead databases and automation tools report contacts and messages. The team wants to know one thing: did a person want to talk about buying?
• Each campaign starts from zero. Nobody records who accepted, who replied or who bought.

What Deberli is

Deberli is an outbound engine. It uses the customer's own LinkedIn account and mailbox, and it sends one sequence on both channels. The customer keeps the relationship with each lead. Deberli does the work between the first search and the reply.

How it works

The customer acts only at the last box. The other steps run automatically. Each reply, positive or negative, goes back into the next search.

The one number

We measure Deberli with buying conversations per person contacted. Reach is limited. LinkedIn limits the number of invites each week, and each mailbox has a safe number of emails each day. Each feature must turn more of these contacts into buying conversations.

Ads to the same people

A lead sees the company name in a message. Then the lead sees the same name in an ad. This makes a reply more likely. Ads are the next channel in the sequence. This feature is not live yet.

• On Meta and Google, Deberli keeps one audience for each customer. The audience contains only the leads who accepted the invite or replied. Each day, Deberli adds new leads and removes the leads who said no or asked to stop.
• A follow up message can contain a link to the customer's website. The customer's pixel records the visit, and the ads show to these visitors.
• On ChatGPT, the ads show in conversations about the customer's topic, to the same type of buyer. This is possible only where OpenAI permits it.

The customer uses their own ad account and pays the platform directly. Deberli makes the audience and updates it. Deberli never uploads cold leads.

What the data becomes

Each campaign makes a record: who Deberli found, why, which messages and ads the person saw, and what the person did. This record grows with each customer. In time, it tells us who will reply and who will buy before we send the first message. A new competitor does not have this data.

We build the predictions in this order. Each step uses the data from the step before it.

1. Who will accept. Deberli predicts the chance that each lead accepts the invite. The weekly invites then go to the people who will most likely say yes.
2. Who will reply. Deberli sorts leads by the chance of a reply.
3. Who will reply with interest. Deberli separates real interest from a polite no.
4. Who will buy. Deberli predicts a meeting or a deal. Customers pay for this result.
5. Which message and which channel. Deberli selects the first message, the follow up and the channel for each audience. For example, it decides if an ad must show before the invite.

The predictions need data. We record this data from today:

• The result of each thread: a meeting, a deal or nothing.
• Tests in the same campaign: two messages or two senders, with leads split at random.
• The last LinkedIn activity of each lead.
• The ads that each lead saw before the invite. This shows if an ad increases the acceptance rate.
• Campaigns from many customers.

Principles

1. Signals first, then scale. Leads with a live signal get the first invites. Leads who fit the ICP fill the rest of the weekly limit, so a campaign does not stop when there are few signals.
2. Deberli shows the contacts, messages and replies of each campaign. We judge a campaign by its buying conversations.
3. Messages use the sender's voice and the recipient's language. Deberli never sends from an account without the owner's permission.
4. Sending stays in the safe limits of each channel. Nothing sends until the customer starts it.
5. Ad audiences will contain only the leads who accepted or replied. A lead who says no leaves every audience.
6. We test a change on real campaigns before we say that it works.

Roadmap

Each phase starts only after we pass the gate before it. We start to record results now, at the same time as phase one. Each week without results makes the predictions later.

More from the blog · Start the free trial