Sales AI Email Coach Lavender rolled its V3 release into beta. The full release, expected later this summer, sports faster writing times through improved personalization tools and insights, and an enhanced coaching system with segmented and custom dynamic models.
By design, Lavender doesn’t want to replace sales reps but make them more effective, a sharp contrast with many SalesTech vendors. Lavender’s AI helps sales orgs understand why emails work and uses this understanding to help reps quickly write effective emails.
“Everything that we do is steeped in this concept of enablement and trying to get the user to think. There’s nothing I hate more than the concept of AI taking thinking away from sellers, because that’s where their growth is going to happen. The friction that we introduce is always intentional.”
COO Will Allred to GZ Consulting
The objective is to educate the sales rep on best practices that raise response rates and move deals forward. Blackboxing these practices doesn’t educate the rep and it ignores the wisdom of the rep.
While Lavender speeds up the composition process (efficiency), its true benefit comes in improving message quality through personalization and email best practices (as determined by AI analysis of client email data). Simply pushing out more quasi-personalized emails provides limited lift.
“If you actually customize emails to a one-to-one message, we’re seeing companies not only getting 25% plus response rates to their cold emails, but they’re generating a majority of their pipeline via those emails,” said Allred. “For example, a Forbes ‘Cloud 100’ company is seeing about 1900% more pipeline delivered via these emails.”
Sales reps struggle to provide “thoughtful personalization” as they need to search across their email, CRM, Gong conversations, and Sales Intelligence solutions. Reviewing and synthesizing this content can be quite time consuming. Furthermore, email platforms lack analytics, so are a “giant black box” when it comes to determining which emails are effective.
Sales reps can select which context sources to employ when drafting an email, with Lavender recommending potential icebreakers and then generating the message.
Lavender addresses the issue of analyzing personalized emails, which creates a “complete blind spot” for Sales Ops. If each email is unique, RevOps can no longer perform A/B analysis of templates.
“In our dashboard, we’re helping managers understand exactly why reps are having success when it comes to email because it’s not an A/B comparison,” argued Allred. “There are hundreds of emails going out. They’ve all been personalized and customized. There are hundreds of variables that are at play here, and so we’re measuring all of them and pulling out what’s most important.”
Lavender’s email coach evaluates your message real time against your historical best practices to recommend changes (brevity, short sentences, non-spammy headers, questions) that raise response rates (2-3x on average) and continue deal progression.
Allred remarked that cold emails shouldn’t get all of the attention, arguing that the key to moving companies through various stages of the funnel is raising response rates. Unfortunately, one of the reasons that email threads die is because sales reps stop asking questions. So, Lavender builds in personalization and encourages questions in responses.
“If you actually customize emails to a one-to-one message, not only are the emails getting 25% plus percent response rates, but the pipeline that you’re driving from that is almost 2,000% more per email.”
Lavender analyzes data across the inbox (Outlook, Gmail) and SEP (Salesloft, Gong Engage, Outreach, Groove, Apollo) to create custom scoring recommendations. It also marries CRM (Salesforce, HubSpot), Conversational Sales (Gong) and third-party data to assist with drafting personalized emails with high response rates.
The AI can examine the emails and determine which phrasing, formats, styles, and wording are effective and which are falling flat. Lavender is diving deeper into AI analytics, looking to discern which emails are effective and why.
“Our approach to this [problem] is to stitch together an understanding of what’s happening in both inboxes, pull that together into a single pane of understanding, and then feed that back to the rep in real-time to give them the coaching they need. We pull all of that information to the forefront for the user.”
It’s a pure understanding of why an email works from an activity standpoint just like the metadata that we capture today, to the persona analytics, to the writing style data that’s been at the heart of Lavender since the beginning. Not just the style but the content itself.”
Lavender CEO Will Allred
The new release will also introduce a set of open-sourced email frameworks that help sellers tune their logic based on the scenario. For example, if a seller is trying to write an email to a prior closed lost opportunity, a framework would help the seller think through the message and subsequent follow-ups.
“We’re going to surface frameworks to help people think through better ways to write the message,” explained COO Will Allred to GZ Consulting. “We want our users to become better writers and better sellers. One of the things we found helpful in creating that was open-sourcing and bringing our frameworks to the surface so people could understand the logic behind what makes a great email.”
Lavender also added a ChatGPT style modal that allows the user flexibility to brainstorm, receive writing advice, or research via searching (or prompting) the web.
Lavender launched its service before the Cambrian Explosion of GenAI apps and subsequently built an “LLM stack” that includes its own models, Anthropic, OpenAI, Watson X, and more. This early mover advantage allowed it to build a solid customer base. According to the Chrome Store, Lavender has over 35,000 active users, a scale that exceeds its competitors.
“We have more daily active users than anyone in our space has downloads of their like products,” stated Allred.
