December 5, 2020

Boost Inside Sales Teams Through AI

Artificial Intelligence
Inside Sales
According to Thomas Baumgartner, Sales & Channel senior partner at Mckinsey, "companies that have pioneered the use of AI in sales rave about the impact, which includes an increase in leads and appointments of more than 50 percent, cost reductions of 40 to 60 percent, and call-time reductions of 60 to 70 percent. Add to that the value created by having human reps spend more of their time closing deals, and the appeal of AI becomes even clearer".

Now that AI use cases are reaching the fields of language understanding, speech to text and emotional intelligence, one could easily wonder where this technological shift is bringing these AI pioneer companies.

What are Inside Sales Representatives challenges and how could AI help ?

Inside sales challenges

While B2C interactions are progressively shifting towards digital (chatbots, rating, IT assistance, ...), the need for human touch in B2B sales has never been so vital. Building long-lasting partnerships and trustful business relationships allows businesses to look to the future with confidence.

B2B client expectations are nonetheless more and more demanding, individuals being influenced by personal customer experiences. ISR leaders currently face several challenges :

  1. Turnover : average ISR turnover rate is above 30% per year (Bridge Group Report).
  2. Recruitment : the majority of IS leaders indicates that they hire ISR with only a 1 to 2 years experience, as skilled representatives are difficult to find (BAO Report).
  3. Training : half of companies spend less than 3 weeks to train new ISR (BAO Report). This represents a real issue for businesses, and especially for companies selling a wide range of different products or very complex services.
  4. Expectations : sales leaders live through one of the biggest changes in the way business to business customers engage in the buying process. As basic needs (product documentation & technical specifications) can be addressed online, customers expectations rose.
  5. Call Time Spent : only 33% of ISR time is spent selling to a client (CSO Insights) as ISR are also responsible for filling CRM's information and client documentation.

AI features

Natural Language Processing (NLP) refers to the use of language-based AI models and addresses interactions between computers and human (natural) languages. Machine Learning algorithms are used to apply grammatical rules to group of words and derive meaning from them. Models and use cases are often based on various AI syntax techniques and rely on complex training phases requiring large amounts of data.

Current NLP algorithms mainly rely on semi-structured and unstructured text data. However, the recent exploitation of raw audio data is opening a brand new AI use cases field. Speech-to-Text methods can also be used to structure audio data through specific learning techniques.

Now, imagine how useful it could be to know that the customer who's calling is "Frank Dart", and dynamically have a pop-up coming to your screen indicating all current and past cases that the company had with him. Or that he is “angry” and that he has just mentioned one of your competitor's name or product name. During a client phone call, how convenient would it be for an inexperienced ISR to be able to consult quoted products and various product specifications without having to search through all the company documentation ?

Different Artificial Intelligence state-of-the-art features can thereby drastically improve your Inside Sales' everyday efficiency :

  1. Keyword / Topic Spotting (KWS) : technique that allows words to be identified in a recording or in a conversation that is underway. You can identically identify topics inside a conversation, whether it is achieved through KWS or different learning training techniques.
  2. Emotional Intelligence : individual’s capacity for identifying and understanding their own emotions and those of the people around them. It's also the ability to use this information to guide thinking and behavior. The capacity to enhance emotional intelligence based on voice analysis can gives ISR extra insights on how to manage the call and reassure their interlocutors.
  3. Next Best Question Recommendation : feature aimed to be fully integrated in the existing sales process. Allows ISR not to forget the selling strategy the company defined, nor basic need discovering questions he could have forgotten. This feature is even more efficient for inexperienced individuals, who can follow AI questions advices all the way.
  4. CRM Output Auto-fill : inside sales representatives actually spend a lot of time filling the CRM and completing post-call activities when they should be calling prospects or clients. AI can help reducing this activities by capturing information in real time during the call, and filling up the CRM forms for the representatives.
  5. Similar Products & Upsales : feature already implemented in widely used services such as Amazon or Youtube with product and video recommendations. With a live AI solution, one could imagine help ISR in their selling process, dynamically informing them of AI recommendations such as "Frequently bought together products" or "Similar products"as those that are mentioned.
  6. Comparison to Successful Calls : the learning process of Sales Representatives can widely be improved using AI analysis on recorded calls. It allows professionals to spot best practices such as : what questions have been asked, how long the sales professional spoke in comparison with his client, which features clients found interesting, etc.
  7. Who to call : feature allowing sales managers to identify high priority leads and/or actions, accordingly to clients' propensity to buy and depending on the process stage in the sales cycle. Such a functionality assists the management to plan effective sales campaigns and focus selling teams on likely revenues.

Business Implementation

For businesses with Inside Sales Representatives, the main question is : should you invest in developing such AI models or should you outsource it ?

Competencies & Technical Skills

The expertise to design and build a full-blown AI platform in-house and to train AI models are tremendously rare and expensive, depending on the type of solution your are looking for.

Data scientists capable to build such architectures also require specific equipment, capabilities and management resources. Product Owners as well as a CTO with technical knowledge could also make a significant difference, depending on how complex your AI project is.

Solution Quality & Continuous Improvement

Before answering the "buy or build" interrogation, you need to ask yourself the following questions : how much efforts can my company afford to put in this project ? What quality can we expect to deliver ? Will it be enough for our customers ? How can we plan to maintain and improve it over time ? AI solutions demand colossal investments and continuous improvement to work at full potential. Maintaining models up-to-date can be really painful, as your teams will need to implement new methods and read new papers to keep your customers' experience at the top level.

Building an AI-Ready Organizational Architecture

An alternative way for businesses could be to be prepared to use AI and data-driven solutions in acquiring specific individual skills and in building the right IT architectural components. For businesses, it would imply recruiting model implementation experts, defining data flowcharts and AI training processes.

  • Implementation skills

While your company doesn’t need the expertise to create elaborate Neural Networks or complex AI models, an AI-Ready organization should hire and maintain a core team of experts capable to customize AI solutions as well as make it fit to the needs of your company. Moreover, building an AI vision, identifying business-driven use cases and highlighting success stories are key concepts for setting-up such an organization.

  • Artificial Intelligence Model Training

An AI-Ready organization needs to be technically prepared to implement AI tools. To that purpose, businesses should adequately create, collect, store and analyze data so that AI solutions have the best ground to flourish when required. The capacity to design data flows and data sets feeding paths should thereby fall on your workforce' shoulder.

In the end, being ready for AI solutions to be implemented doesn't mean everything is about AI. The power of this "AI-Ready" alternative is that it allows you to stay focused on your customers and on their experience while succinctly infusing AI and data-driven solutions when required.

Call to action : CogNeed AI™

Many companies are positioned on the "AI for sales" segment, either adding features including various levels of AI to an existing application or building new solutions from scratch.

At Cogneed, we focus in developing AI apps specifically trained for your team, to assist the remote sellers in real-time, to accelerate ramp-up, have better conversations, provide more value to clients and increase productivity.

If you want to hear more about our solutions, please visit our website at, or contact me at