AI In Sales: What Do The Experts Have To Say? (1/4)

How to effectively use AI in the sale process, to boost lead volume, close sales, conversation intelligence, effective coaching, and more?

Initium AI has recently hosted a discussion on the role that AI can play in sales, with sales and AI experts who shared their insights and experience:

Raluca Banea , Director of Strategic Accounts & Innovation at MedEquip Depot

Alex Gohl , Senior Director of Client Development at Syndigo

Rada Mihalcea , Professor & AI Lab Director at U. Michigan and Co-founder of Initium.AI

We are excited to share the first part of this conversation.

Highlighted here:

• Basic terminology in AI
• AI tools, applications and systems to be aware of in sales

Raluca : Welcome to today's webinar. We're really excited to be here with you today, talking about the very important and timely topic of the use of AI in various aspects of the sales process. I'd just like to start quickly by defining some key terms to ensure we are all on the same page. First, let's talk about artificial intelligence (AI).

In simple terms, AI involves using computers to perform tasks that traditionally require human intelligence. It allows computers to process large amounts of data in ways that humans cannot, enabling them to recognize patterns, make decisions, and judge like humans. Machine learning is a subset of AI and refers to the capability of a machine to imitate intelligent human behavior. It's fascinating that computers can learn and adapt over time without explicit instructions, using algorithms and statistical models to analyze data and infer patterns.

Another important term is Natural Language Processing (NLP), which refers to the ability of a computer to interpret and understand human language. Once we have these terms defined, I'd like to briefly discuss the inner workings of current AI systems. Rada, could you please provide some insights?

Rada : Certainly. In recent years, there have been significant advancements in deep learning, which has contributed to the progress we see today. Deep learning is not a new technology but rather an advancement in neural networks, which have been around since the 1950s. The significant change now is the availability of more computational power and a wealth of data, including web data used to train these systems. Neural networks aim to emulate the functioning of the brain. At their core, they consist of neurons that receive inputs, perform mathematical operations on them (often with weights), and apply functions to produce outputs. The functioning of a neuron is actually quite simple; it just involves mathematical operations, a linear combination of its inputs, followed by a simple non-linear transformation. The power of neural networks lies in the weights on the connections between neurons, often referred to as parameters. Large companies often boast about the billions of parameters they have.

Another noteworthy aspect is the prevalence of large language models in NLP. These models predict what follows based on the context of previous statements. When trained on vast amounts of data, they can generate output like auto-complete suggestions. Models like ChatGPT, for example, learn from a tremendous amount of data, including web data consisting of around 400 billion words, far surpassing what an individual can read and accumulate in a lifetime. Additionally, there's input and supervision from humans to ensure the accuracy of the models. This combination of data, code, and human input contributes to what we see in today's language models.

Raluca : I don't think there is a better time to have this conversation. Now that digital transformation has really accelerated over the last three years or so, and even yesterday Salesforce unveiled Einstein GPT, which is a generative AI tool that they now use across all of their Salesforce clouds to generate personalized content for different clouds. AI technologies are moving at a very fast pace, and we can definitely expect huge changes to come in all areas that we work in. But today, of course, we are focusing specifically on sales.

AI is already being used to automate and streamline many of the sales processes, such as lead generation, upselling, cross-selling, customer segmentation, personalized sales pitches, forecasting, and even setting appropriate discounts to try to increase the likelihood of winning a transaction. Rada, could you share with us what types of AI tools, applications, systems, and companies we should be aware of? And Alex, if you could give us an example of a practical setup of an AI pipeline of applications that is using the sales process, that would be really useful.

Rada : There are some broad areas that I think of in terms of where AI is already making a difference and can make a difference in the sales pipeline. Keeping all the relevant information in one place and then processing it in smart ways, like figuring out when is the right time to follow up on an email or drawing connections between relevant information for an account.

In addition to Salesforce, there are other companies in this space like Zoho, HubSpot, SalesNow, Follow Up, and several others. Another broad area is customer communication, and of course, there are different ways to communicate with customers. Some that we've seen are in customer emails, where you can write templates or trigger follow-ups, and here there would be tools like Seamless AI, Conversica, and several others which are focused primarily on this space.

Transcription, for instance, is used by many of these tools, and transcription is an AI product that does automatic speech transcription. Here we see other tools like Nuance, iSpeech, or DeepTalk, which essentially take a conversation like ours and transcribe it, so then you can also have access to the language, not only the speech. On top of that, you can also have smarter processing of what was in a conversation. For instance, finding some of the topics, and we see tools like keyword search. I believe Gong and Chorus are big players in this space of meeting transcripts and being able to search for keywords or an issue. We also have the ability of processing conversations, adding additional layers like summarization of a conversation, like note-taking.

Also, conversation analytics and customer feedback, including the analysis of the sentiment in a conversation, or the empathy and the questions being asked. Or to identify what products a customer enjoys the most, which could be used for maybe advertisement on growing those areas. Or figure out what are the pain points as a way of finding what needs to be addressed and also as a way of finding white space.

I believe training is also a big area. There are different parts that one can use from conversations to provide feedback to sales agents, so they can learn from previous mistakes and improve.

Alex : As far as what kind of setup you might find or solutions that you might find with a business, I think it really depends on what kind of resources they have available, and what are the functions that are appropriate for a given point in time. For instance, a large enterprise is going to need very deep robust tools, and they'll probably need all of them, whereas a growing business may spend more time figuring out what to choose from among all the tech that's available.

There's just so much information out there, and a lot of that admin stuff where reps are just going out, and if they're a public company, there's information they can look through like annual reviews or companies' websites to get all that information together. It takes a lot of time, and so I think solutions like ZoomInfo or Seamless AI to find who are the right people and businesses to talk with and when and how to contact them are really important. We're seeing a lot of reduction in admin work for sales reps from AI tools.

It's also really exciting to see some of that predictive information that you kind of talked about, the different levels of intelligence that power these systems. In sales, we have a very complex set of problems that we're solving for customers, and across a lot of different teams. We work with marketing, sales, IT, e-commerce teams, and there are a lot of different folks with different priorities and problems that we're solving. So it's essential to tie all this back to sales enablement or readiness platforms, whether we're helping and bringing in new reps and trying to train them about what we do, or upskilling existing folks as things change very fast. There are solutions out there that we're implementing to basically take all of our content that we use to help people learn. Putting these AI tools together with all the content we have provides great coaching opportunities for us as managers or sales leaders, and helps us be more prescriptive.

So just to recap, business intelligence, enablement or sales readiness, conversation intelligence, decision tools that help predict and surface up information to sales leaders so we can make decisions on where to spend our time and close deals, or predict what's likely to close and what's not.

We also see forecast intelligence or pipeline management intelligence, with analytics and information to help coaching and reps learn and prioritize their deals. It's really fun to see those kinds of advancements so we can be smarter as businesses.