How AI can help telecom companies maximize SMB sales

AI is set to equip telecom companies with the intelligence and precision they need to grow their fiber market share in this important, but scattered, market.

Small and medium-sized businesses are a lucrative market for telecom companies. After all, signing up a new customer for fiber optic services means multi-year revenue and low maintenance costs. Yet SMBs are a fragmented market to sell to. They span everything from two-person shops to 200-employee businesses. Artificial intelligence is a natural fit for growing SMB market share and reducing the sector’s high acquisition costs—but only when used in the right ways.

Small business, big challenges

Depending on the source, SMBs represent anywhere from 15% to 40% of total revenue for telecom companies. To the multi-tasking owners and decision makers for SMBs, however, coordinating telecom services—phone, internet, digital tools and cloud services—is just one of many responsibilities they juggle daily as part of their job.

The painful reality for telecoms is that no SMB owner is waiting for a call about connectivity services. Yet the market for SMB services has become hotly competitive. With the proliferation of private equity-backed fiber companies, driven partly by government subsidies for broadband expansion, the number of fiber providers in most markets has significantly increased.

Worse, small businesses take many shapes, making it difficult to predict their connectivity needs. For example, is the home-based establishment a growing enterprise in need of high-speed internet and fixed wireless access, or is it a side hustle, such as a photography lover selling seascape images through an Etsy shop? What’s more, remote work and hybrid models raise thorny issues for telecoms when it comes to sales commissions. If a company pays for services such as VPN for remote workers, does the commission go to SMB or residential sales?

The winding path to the right buyer

To crack the market and reach SMBs, telecoms pursue a variety of channels, such as direct mail, email and targeted advertising. But the largest spend is on outbound voice. It remains the only channel through which companies can be reasonably sure of getting a response. Yet research we conducted for outbound voice calls indicates that reaching the right buyer—a decision maker who has budget—is akin to finding a needle in a haystack. The average pickup rate among SMBs is a scant 2%. Just 1% of calls connect with the target customer. The close rate is a tiny .05%. The final tally? Two hundred calls are required to close one deal.

AI offers telecoms a hedge against the steep challenges of selling into the SMB market—but not in the ways you might expect. True, generative AI is there to craft custom value propositions. But long before telecoms make their pitch, AI can wield the laser precision they need to reach the right person. How can you pinpoint the SMB that has the internal spend approved to, say, upgrade from 1G to 2G Internet?

Here are three areas we’re exploring where AI has the potential to have a major impact:

• Identify the optimal way to access SMBs. Successfully selling new services to SMBs means meeting business owners on their terms. Are they more likely to respond by phone, or through digital channels like email? Do they prefer a direct message via social media? Is an in-person contact the best choice? This insight is a critical first step; all the data in the world won’t help you if you can’t talk to prospects. AI can help predict the best method of outreach. Because AI constantly learns and correlates successes and failures with channels over time, it can identify the best channel for a specific SMB customer. The correlations could be based on variables such as vertical segments, company size, geography and ownership structure.
  
  In addition, AI has the potential to suggest access based on vertical subsegments. That is, drilling down to credit unions, and then to branches located within certain city blocks that service teachers and postal workers. We’re exploring how building that level of intelligence customization will ultimately help telecoms become more effective in SMB outreach.
  
  
• Automate calls to eliminate the tell-tale pause. For recipients, the pause that occurs as predictive dialers hand off a call to a live agent is a dead giveaway that they’re about to be pitched. We’re working on a model that automates the calls and makes data immediately available to intelligent agents—bypassing the need for today’s detect-and-transfer experience as live agents wait for their screen display system to populate with relevant offers and discounts. With AI agents, as soon as the prospect picks up, the AI agent has the data and can make the relevant offer: “Save 50% for six months on the cost of a monthly subscription. Press 1 to sign up. Press 2 to speak with an agent.”
  
  
• Using AI to speak in local dialects. Although the first few seconds are make-or-break for any outbound call, the timing is especially critical for win-back campaigns. A customer who has cancelled a subscription or service remembers your brand, possibly with mixed feelings. Introducing a comeback offer is also about rebuilding trust. To help telecoms put their best foot forward, we’re exploring agentic AI’s potential to open outbound calls with a voice that replicates local dialect. Instead of the dreaded hello-how-are-you, agentic AI is ready with a custom opener that sounds and feels natural, like speaking in the owner’s native language, adopting a regional accent (we hear you, Boston), or using a conversational pattern that’s characteristic of Gen Z (“This is literally the best offer ever.”).

In today’s cutthroat fiber market, AI isn’t just a tool—it’s the sales commander SMB leaders need. By pinpointing the right prospects, optimizing outreach channels and delivering hyper-personalized pitches, AI transforms the uphill battle for SMB fiber customers into a strategic, data-driven campaign. The winners will be the telecoms that deploy AI not as a gimmick, but as a tactical advantage.