Contact centre and commercial upsell: how to connect them without toxic selling

An honesty test

A simple test you can run inside the organisation. Take a random call recording in which a customer called about a problem and during the call the agent offered a tariff upgrade or an additional service. Listen as the customer. Answer one question: did the agent offer this because it really fits the customer, or because the agent has a sales target?

If hesitation is needed, the model the contact centre is sitting on is probably already broken. And the customer hears it, even when the agent is doing their best to sound natural.

This article is about how to connect care and commerce in the contact centre so that upsell is not toxic and at the same time delivers commercial results.

Where the care-commerce conflict comes from

In most telecoms the curve evolved this way. First the contact centre was pure customer care: take the complaint, solve the problem, say goodbye. That worked while connectivity margins were high.

With market saturation and falling connectivity margin, monetisation pressure appeared. The contact centre was a high-contact channel — the customer was already on the line, the conversation was already happening. The logical move was to add upsell. A sales KPI was introduced.

The path from there is typical. Agents start sounding like sellers, not helpers. The “customer listens to the offer through” rate falls. Complaints about pushy selling rise. At the same time the care team starts to see upsell as “extra load that gets in the way of solving the problem”. The sales KPI delivers a number in the report, but real margin is questionable — conversion is low, post-sale churn is high.

In the end the team splits: some run “pure care” (and miss upsell), others push selling (and lose trust). Neither model is optimal.

Today — four observed patterns

Across operators today, four approaches to this conflict coexist.

Pure care without upsell. The contact centre solves problems only. Upsell is a separate outbound channel and does not overlap. Upside: care stays clean, NPS is high. Downside: a missed commercial moment when contact has already happened.

Aggressive upsell on everyone. Every call ends with an offer. Conversion in the moment can look respectable, reports look positive, but post-sale churn is above average and NPS is dented. Annualised financial effect is often negative, but harder to see in a weekly KPI.

Mix without rules. Sometimes upsell happens, sometimes not. The decision is left to the agent. Behaviour is uneven, training is hard, audit impossible. The customer does not know what to expect.

Contextual upsell with rules. Upsell happens only when explicit conditions hold — the problem is solved, the customer is in a calm state, there is a concrete signal (data-based) that the offer is relevant. The hardest approach to launch and the most productive.

Most telecoms today sit in approach 2 or 3. Those who moved to 4 stand out on retention and NPS over the long horizon.

What changes after 2026

Biometric requirements for digital banking from April 2026 (cbu.uz) indirectly affect the contact centre. Operations that used to flow through the app or self-service now require verification that pushes the customer to a live agent. Inbound mix shifts — more contacts about “could not pass verification”, “cannot log in”, “need to change device”.

These contacts are structurally not where upsell is appropriate. The customer is frustrated, has an access problem, and an offer to upgrade the tariff at that moment looks like a mockery.

In parallel, customer expectations of meaningful contact rise. The customer already sees in banking, e-commerce and large digital services that communication is contextual. If telecom calls with an offer that ignores yesterday’s complaint, trust takes a hit.

In these conditions the third and fourth approaches become dominant. The first two are legacy, and operators not rebuilding will lose position.

What is needed for upsell to stop being toxic

Suppression rules before upsell. If the customer is in an active billing dispute, in a recent complaint, with an open case — no upsell. An absolute rule, automated, not left to the agent’s intuition.

A contextual signal that the offer is relevant. Not a generic “we have a package — sell it”, but a signal from data: “this customer has a roaming pattern to Turkey, the international package fits”. Without a signal — no offer.

Timing within the call. Solve the problem first. Only when the customer has visibly entered a neutral state does the offer come up. The transition is natural — “while I look at your account I noticed you fit…”.

The ability to say no without pressure. The customer says “not now” — the topic closes. No second pitch, no “but think about it”. This changes the psychology of the dialogue.

The care team’s KPI on retention and NPS, not just sales. With a pure conversion KPI, behaviour always slides into aggression. A balanced KPI structure — sales plus retention plus NPS — produces different discipline.

What to prepare now

Audit of the current model. A random sample of 50-100 calls with upsell. Listen with a critical ear. Allocate to the four patterns. That gives an honest baseline.

Suppression rules in the contact-centre operations system. Which signals automatically block an upsell attempt in the current conversation. The most modest move with fast effect — it does not “improve sales” but it stops token churn.

The contextual signal. Build a minimal data pipeline into the contact centre that gives the agent at the moment of the call not just customer data but 1-2 recommendations: “this fits now, here is why”. An investment, but it pays back through higher in-the-moment conversion.

KPI rebuild. Move from pure conversion to a mixed metric. The sales head will resist — a sales number simplifies the picture. But without this move tactical fixes do not help.

Tone training. Not “how to sell the upgrade”, but “how to offer when appropriate and how to accept a refusal without pressure”. A different skill set.

The rebuild period is 6-12 months depending on contact-centre size and operations maturity.

When not to launch the rebuild

If retention is fundamentally low for other reasons (network, billing, brand), upsell is not the priority. Core problems first, monetisation second.

If the contact centre is outsourced and the contract is rigidly tied to a conversion KPI, the rebuild hits the commercial contract. Without renegotiation, the third party will not engage.

If CRM does not show the agent the customer’s context in real time (recent interactions, open cases, complaints), suppression rules cannot work technically.

If sales leadership is not ready for a temporary dip in conversion in the short term, the rebuild loses cover. Pressure to revert in 1-2 quarters will be high.

If leadership does not agree that long-term retention beats short-term conversion, the rebuild gets de-prioritised continuously.

Discussion points for the committee

Do we listen to call recordings, and what do we hear? If “we have not for a while”, that is the first step.

What is the current KPI structure of the contact-centre team? And what is the real upsell margin after accounting for churn?

How many contacts a week receive an upsell offer that should never have reached them (open complaint, active dispute)? If unknown — diagnostic.

What do NPS readings show for the segment that received upsell during a problem versus the rest?

Is the organisation ready for a 6-12 month transition with KPI-shifting complexity? Without that, the change stays a declaration.

How SamaraliSoft can help

Contact Center Care-Commerce Realignment — audit of the current model through call sampling, redesign of suppression rules and contextual triggers, restructure of team KPIs, design of a real-time contextual signal data pipeline, and a 90-120 day pilot on one team measuring retention, NPS and conversion together.

Related reading

• /en/insights/telecom-nba/ — NBA instead of mass campaigns
• /en/insights/telecom-subscriber-intelligence-operating-model/ — operating model
• /en/use-cases/telecom-churn-war-room-mnp/ — retention in the MNP era
• /en/insights/telecom-arpu-bundles-devices/ — where ARPU grows

Sources

• Central Bank of Uzbekistan — biometric verification for digital banking
• Digital Government Portal — digital infrastructure UZ
• UzDaily — Central Asia telecom market
• Ministry of Digital Technologies — telecom in UZ