How SMBs can use machine learning to kick ass

Being a small or medium business (SMB) is not easy. They are often forced to stand against jumbo-sized opponents – enormous international corporations with budget to spare and people to assign to any chore imaginable. But SMBs have one major advantage that large corporations cannot replicate – agility. Team training, adoption of new tools and organisational changes can occur faster with less friction.

Being compact means that you can start using new technology faster, and your weapon of choice should be machine learning.

Why SMBs should consider machine learning

Machine learning is the cornerstone method for ‘teaching’ a machine to act like a human; drawing conclusions based on provided data, spotting trends, making predictions and analysing enormous datasets in a matter of minutes.

You have already dealt with machine learning if you have ever asked Siri to recommend a local restaurant or Amazon to suggest some new books to read. Both companies deploy advanced machine learning algorithms to learn more about their customers and pitch better product offers. And, as a small company, you can do the same without having Amazon’s marketing budgets or huge development teams on board.

Personal recommendations on the Amazon website as an example of machine learning in e-commerce
Amazon uses machine learning to enhance personalisation

Machine learning tools and ready-to-use proprietary platforms can now accommodate an SME budget and generate significant ROI within just a few months. So, the next time a larger competitor is trying to kick you out of the league, why not consider these use cases of machine learning below:

Improve customer retention and engagement without hiring a bigger team

How do you weed out the good leads and the most prospective customers? Probably, by looking at data stashed inside your CRM or a bunch of spreaders. And how long does it take for a sales rep to go through all those figures and make some sense out of it? Probably longer than the average customer is willing to wait on the line.

For a small business, it is crucial to engage and activate every good lead coming your way and stay in touch with long-term customers. However, due to limited team sizes, it may be hard to keep tabs on all things at once. So what happens if you send a machine to score incoming leads and predict the probability of a sale?

Alvin Electronics, a high-performing team of eight people, decided to test that idea. The company was struggling to reduce customer churn. As everyone already felt swamped, the sales team had no time to manage and engage with customers who were about to walk away. So they decided to give machine learning a try:

  • The platform they used assigned a churn probability score to each customer in the CRM. The top-20 list was sent immediately to the sales team, who took preemptive action. As a result, the company managed to achieve a 70% retention rate.
  • During the first month, Alvin Electronics managed to increase their revenue by nearly $9,000 by re-engaging with existing customers.
  • Machine learning tools also helped them to pindown the average transaction value per client, average time between transactions and the actual current churn rate. This way the sales team could immediately spot new opportunities and take action to activate some customers.

Create highly-personalised content marketing campaigns

It’s no secret that outranking a household-name competitor in search results may seem like mission impossible for a smaller provider. But that doesn’t mean you should exclude blogging and content marketing from your SEO strategy.

Maybe you won’t have the ‘weight’ to get straight ahead of a larger company for highly-commercial search terms. But an intimate rapport with a select group of customers can become your unique strength.

By the end of 2018 companies that have incorporated personalisation strategies across multiple channels (web/mobile) will outsell businesses that have not by 20%.  Machine learning algorithms and ready-to-use tools can help you create highly-personalised email marketing campaigns, custom tailored landing pages and develop compelling content that captivates different customer segments.

Quotation that by 2018, organizations that have fully invested in all types of personalization will outsell companies that have not by 20%

The best part is that you can always experiment with ‘light’, budget-friendly personalisation solutions before developing a custom machine learning algorithm:

  • Try Optimizely X to create dynamic landing pages. The tool will adjust the page layout and offers based on the customers’ past actions, affinity with a brand or other factors. For instance, you can program it to show new products or seasonal discounts to returning visitors.
  • Our proprietary software, Apollo Insights, can crunch the search marketing data for you and help you understand what kind of content is resonating with your audience the most. It can suggest what type of content to create for specific goals, and help you identify new content marketing and SEO opportunities that your competition may have missed.
  • Use quizzes to solicit more personal data from prospects. Fidelity, for instance, asks new customers to complete a simple financial checkup. The quiz helps determine the person’s preferences and intent. Based on provided responses, the company then creates a personalised content digest. This way, Fidelity gathers more data about the customer that could be later leveraged during the sales process, whereas the customers receive content tailored to their financial situation.
Question about spending used by Fidelity to build up a customer profile and personalise their website's user experience
Fidelity uses questionnaires to help improve its personalisation

Streamline customer support, pre-sales talks and on-boarding with chatbots

As machine learning techniques keep advancing, chatbots are becoming capable of handling more and more complex chores – from providing basic information and directions to a company’s office, to making hotel reservations, scheduling meetings and resolving customer support issues.

Perhaps, if hiring more people in one of your teams isn’t an option, opt for a chatbot to handle some mundane tasks. Juniper Research predicts that by 2022, chatbots will save businesses more than $8 billion annually by automating low-value processes.

Chatbots have also proven to be effective in driving profits. SnapTravel is a new hotel booking service with a twist. They operate entirely through Facebook Messenger with most of the customer queries being handled by bots, instead of human agents. In 2017, the company generated over $1 million in revenue on Facebook Messenger alone and has now expanded to other conversational platforms like Viber and WeChat.

Advert for SnapTravel showing use of Facebook Messenger to allow customers to book hotel rooms

However, as the founders noted, SnapTravel’s success would not have been possible without extensive conversational UI personalisation (and we second that!). The app extracts the individual preferences from the conversation  – e.g. if a customer favours luxury properties or prefers free breakfasts – and applies that knowledge to future recommendations to increase conversions.

The travel industry is already a ripe field for data science and machine learning, and other industries should be taking the lead. Small businesses in any niche can send chatbots to handle the initial conversation with a prospect, gather required data and then have some machine learning tools to transform it into meaningful insights and action. This way your small company can become an agile powerhouse – capable of delivering the same, or even superior, service to customers without ‘big player’ budgets or teams.

George Karapalidis profile picture
George Karapalidis

George is an SEO Specialist and Data Scientist in the Portsmouth office. He has worked in E-Commerce and Digital Marketing across many industries, for and with companies all over Europe. Before joining Vertical Leap, George worked as Marketing Director for his own company, for which he managed to expand the company’s activities to 5 European countries. George has been creating websites for more than 10 years and he has in depth experience in designing and bringing optimised E-Commerce websites to market.

More articles by George
Related articles
Traffic lights image illustrating the article Data science for marketers (part 1): Getting your data ready

Data science for marketers (part 1): Getting your data ready

By George Karapalidis
5 ways machine learning should change the way you do SEO

5 ways machine learning should change the way you do SEO

By Chris Pitt
2018: The year of small business SEO

2018: The year of small and medium business SEO

By Kerry Dye