Telecommunications · Upsell Call
The Upsell Script for Telecommunications Customers: Turning an Account Review Into a Second Region
Your customer picked up expecting a check-in. They've got a POI incident from three weeks ago still generating ombudsman complaints, a field services budget that their Director of Field Operations is already defending line by line, and a capex committee that doesn't sit again until the second week of next quarter. They are not sitting there hoping you'll mention a new module. Whoever you're calling — GM Service Assurance, Head of Network Operations, Director of Field Operations, or the GM Consumer who can name the exact suburbs that churned after the last mass service disruption — their default posture is three-layered: the money's committed, the team is mid-migration with zero spare cycles, and privately they aren't sure they're getting full value from what they signed last year.
That last one is the whole call. If you open with "I wanted to walk you through something we launched," you get a polite ten minutes and a "send me something." If you open with their no-fault-found rate in the region where you're actually deployed, split between tickets where your degradation flag fired and tickets where it didn't, you get a real conversation — and a buyer who starts correcting your numbers, which is exactly what you want, because their correction tells you which metric they defend in their monthly pack.
This is an upsell script for Telecommunications customers built around one move: anchor on a result they already own in their own numbers, name the underuse before they do, then reopen the piece of scope they deliberately cut during the original deal. Almost every telco expansion was already scoped once and killed for budget or belief. You're not selling something new. You're re-litigating a decision they made twelve months ago with less evidence than they have now.
The upsell call script
Say it in your own words. The structure is the part that matters.
- 1
0. Pre-call usage audit — do not dial without this
Fill every blank from your own systems before you touch the phone. If you can't, you're running a value-realization call, not an upsell. **Adoption:** Licensed seats vs. weekly active, split by function. Name the dark corners out loud: "NOC has 24 seats and 19 weekly actives, the L1 triage desk has 16 seats and never logged in after week two." Which regions are live on the northbound feed, which were scoped and never cut over. **Depth:** Which screens they actually work — degradation alerts, dispatch pre-check, the upstream/NNI classifier, the cohort view the retention team was supposed to get. Last login of your exec sponsor. If the GM Service Assurance hasn't logged in since the last QBR, that's your opening problem, not your closing one. **Outcome — theirs, not your benchmark:** One number sourced from their data or their own QBR minutes. NFF rate on flagged vs. unflagged tickets. Truck rolls per 1,000 services in the deployed region vs. the control region. Mean time to detect before the customer calls. Have the query, have the date range, have the caveats. **Commercials:** Annual value, renewal date, who signed, whether that person is still there, discount level, and whether an add-on can be co-termed. **Assurance and escalation history:** Any MSD in the last 60 days. Any period where your alerts were muted during a fault. Any open P1. Any complaint spike they'd associate with an event you touched. **Champion status:** Still in the role? Promoted? Quietly moved sideways in the last reorg? Telco org charts move after every migration milestone. The test: can you say one thing about their network they'd be mildly surprised you knew? "Your Southern region ran 6,100 dispatches last quarter and 21% closed no-fault-found" passes. "How's everyone finding the platform?" fails, and it's the exact question this buyer holds against you.
- 2
1. Frame it honestly in the first 30 seconds
"Thanks for the time. Two things I want to cover. First — I pulled your last two quarters out of the platform and there's a pattern in the [Northern] region dispatch data I want to check with you, because I don't fully trust my own read of it. Second, depending on what you say about that pattern, there's a part of this that I think is relevant to [Southern / the migration regions], and if by the end I don't think it is, I'll tell you and we'll stop. Fair?" Sets an agenda, proves you did the work, and pre-commits you to walking away. That last part lowers a telco operator's guard faster than any rapport-building, because they've been pitched AIOps by four vendors who never once said "this might not apply to you."
- 3
2. Anchor on the result they already own
Make them say the current deployment worked, out loud, in their numbers. "In [Northern], last quarter, on tickets where our degradation flag fired before the customer called, no-fault-found closed at [9%]. On tickets where it didn't fire, [24%]. Same techs, same closure codes, same quarter. Does that match what [Director of Field Ops name] is seeing in his monthly pack, or am I reading it optimistically?" Then shut up. Their correction is worth more than your number. Write it down verbatim — that sentence is the first line of the business case they'll carry to the CFO. Second anchor if the first one gets argued: "What's mean time to detect looking like in that region now versus the rest of the footprint? My read is you're finding [X%] of degradations before the first customer call. Is that the number you'd defend internally?" **If they can't confirm any result, stop the upsell.** "Then let's not talk about anything new today. Let's spend the half hour on why you're paying for something you can't point at a number for." That call saves the renewal. The other version loses it.
- 4
3. Name the underuse before they do
They are already thinking *we barely use what we've got.* Say it first so you're diagnosing instead of defending. "Here's what I'd push back on if I were you. You've got 40 seats. 19 are in weekly and all of them sit in the NOC. The L1 triage desk has 16 seats and hasn't logged in since onboarding. If I were you I'd be asking why I'd spend another dollar before I've used what I bought." Then reframe unused capacity as unrouted work, and be honest if it isn't: "So let me tell you what I think those 16 are doing instead. They're not idle — they're taking the call, running the standard line test, seeing nothing, and booking the appointment window, because that's the only path the script gives them. The work is happening. It's just happening as a dispatch instead of as a triage decision. Whether that's a product gap or an onboarding gap, I'd genuinely rather know before I ask you for anything." If it's an onboarding gap, say so and fix that instead. Conceding the point buys you the right to come back next quarter.
- 5
4. Find the seam, don't pitch the module
Diagnostics, in rough priority order: - "What did you deliberately leave out of scope when we scoped this last year, because it was too big a bite?" — highest-yield question on this call. Most telco upsells are just previously-cut scope. - "When a degradation flag fires and the pattern looks like it's upstream of the NNI, what happens right now? Who decides whether that becomes a wholesale fault or a truck roll?" - "What's the Monday morning fire drill? What's the thing your ops review keeps landing on that nobody's fixed?" - "When [GM Consumer / Retention] wants to know which cohorts were exposed to last month's MSD, how do they get that? Who builds it, and how long after the event?" - "Who's asking you for reporting you can't produce today — CSG compliance, complaint handling, the CFO on cost to serve per service?" You're listening for one specific seam: the moment the network knows something the customer-facing side doesn't, or the moment a dispatch gets booked because nobody could tell whose fault it was.
- 6
5. Build the case in their arithmetic, price included
Unit, volume, capture rate, net. All from their data, all conservative, and say the price in the same breath. "Your pack: [6,100] dispatches last quarter across [Southern], [21%] closed no-fault-found — call it [1,280] vans that found nothing. In [Northern], where we can see the telemetry, just over half of the NFFs had a degradation signature that sat upstream of your NNI. Take half of that, because I'd rather under-promise: [320] avoidable dispatches a quarter. At your fully loaded [$180] a dispatch that's about [$57k] a quarter. The classifier is [$7,400] a month for the three Southern regions. [$22k] a quarter. So roughly 2.5x, before you count the appointment windows you don't blow, the first-time-fix number that moves, and the 'we sent someone and they said it wasn't us' calls that turn into complaints. If you think my capture rate is generous, tell me what number you'd believe and we'll rerun it live." Hand them the pencil. A telco operator who rewrites your assumptions has just adopted your model.
- 7
6. Cost the implementation in hours and names
Bandwidth is the real objection. Budget is the costume it wears. "What I need is two hours from whoever owns the northbound feed to point us at the same publisher we're already reading in [Northern] — no new integration, no data model work, no change window. Then your NOC lead for 30 minutes a week for the first month to set the precision threshold and tell us when it's firing on noise. That's the whole ask. Nothing touches the migration, nothing writes back into your ticket flow until you say so. If it needs more than that from your engineers before you see a number, I've mis-scoped it and I'll tell you." Put everything you can absorb on the table here — your team doing the config, your CS running the NOC enablement — not later as a concession.
- 8
7. The ask: one region, one quarter, co-termed
Smallest unit that proves the case. Success criteria in their words, written down on the call. "Give me [Southern] for one quarter. You define the pass mark — my suggestion is no-fault-found rate on flagged dispatches, plus how many upstream faults we raised to wholesale before a van moved, but if there's a number your exec cares about more, use that one. If we don't hit it, we pull it out and there's nothing further to pay. If we do, we roll it to the other regions at your existing renewal date and co-term it, so you're managing one contract and one negotiation instead of two." Co-terming is the most underused unlock on this call. It turns a new purchase into an amendment and skips a procurement cycle — which in telco is the difference between this quarter and next financial year.
- 9
8. Buying committee check — ask it out loud
"Last time this went through [name] in finance and it was signed off against the assurance line. Is that still the path, or has the threshold moved since [the reorg / the new CFO]?" Follow-ups worth asking on this call rather than discovering in week six: - "If this lands in the field services budget instead of yours, whose conversation is that — yours or [Director of Field Ops]?" - "Is there a bid threshold now? At what number does procurement want three quotes?" - "Does the CFO see cost per truck roll and cost to serve per service in the same pack, or are those two different reviews?" If a new CFO or a new procurement policy has landed, you're running a fresh evaluation dressed as an expansion. Price it and time it accordingly.
- 10
9. Park signals — walk away rather than push
Do not push the expansion if any of these are live. Pushing converts a healthy account into a churn risk to book incremental revenue: - An MSD or major POI incident in the last 60 days with complaints still open - Your alerts were muted or ignored during a recent fault and nobody's debriefed it - Your champion just changed roles in a post-migration reorg - Renewal inside 60 days with an unresolved value question - A flat or declining weekly-active curve you can't explain - A cost-out program or hiring freeze announced this quarter The move when you see one: "Then this isn't the call. Let's use the time on [the muted alerts during the POI event], and I'll come back to the Southern conversation when that's closed out." You lose a quarter. You keep the account.
- 11
10. Recap email, same day
Written for the person your champion forwards it to, not for your CRM. Six things, in this order: 1. The result they confirmed, in their words, with their caveat included 2. The gap they described, quoted 3. The value math with their assumptions, labelled as theirs 4. The price, stated plainly, monthly and annualised 5. The implementation ask in hours and named roles 6. The pilot scope, the success metric they chose, and the decision date One page. No attachments they have to open. If the CFO reads only the subject line and the first three bullets, they should still be able to ask the right question.
How the call actually sounds
Prospect on the left, the rep on the right.
Rep
Thanks for making the time, Priya. Two things. First, I pulled your last two quarters out of the platform and there's a pattern in the Northern dispatch data I want to check with you, because I don't fully trust my own read of it. Second, depending on what you say about that, there's a piece of this I think is relevant to the Southern regions — and if I don't think it is by the end of the call, I'll say so and we'll stop. Fair?
Buyer
Fair, but I'll save you time. Capex is committed through June, the committee doesn't sit again for eleven weeks, and my engineering team is neck-deep in the access migration. If this is a build-out conversation you're a financial year early.
Rep
Understood, and I don't think this is capex — I'll come back to where I think it sits. Start with the pattern. Northern, last quarter: on tickets where our degradation flag fired before the customer called, no-fault-found closed at nine percent. On tickets where it didn't fire, twenty-four. Same techs, same quarter. Does that match what Dave's seeing in his field pack, or am I reading it optimistically?
Buyer
I'd be very careful with that number. We changed closure code policy in October. Techs used to dump everything into 'other' — now they've got a forced-choice list. A chunk of that delta is us measuring properly for the first time, not your tool working.
Rep
That's a fair hit and I'd rather take it from you than from your CFO. Two things. The policy change landed on both cohorts — flagged and unflagged — so it should move the absolute numbers but not the gap between them. That said, if you don't buy NFF as the measure, what's the number you'd actually defend in your monthly review?
Buyer
Truck rolls per thousand services. And Dave will tell you dispatch volume in the North is down, and then he'll tell you it's because we didn't get the storm season we got last year. He's not wrong.
Rep
Then let's take the weather out of it entirely. Different cut: appointments where the technician arrived and the fault sat upstream of your NNI — not yours to fix, van still went. In Northern that's down from about eleven percent of dispatches to six. Storms don't move that ratio, because storms hit both sides of the demarc. Does six-versus-eleven sound like something Dave would recognise?
Buyer
That one he'd recognise. That's the number that makes him swear in ops review. But before you go anywhere with it — I'll tell you what's bothering me. You've had forty seats in there for a year and the L1 desk has never opened it once. I'm not spending more on a platform my own people don't log into.
Rep
You're right and I'd have led with it if you hadn't. Nineteen weekly actives, all NOC, and the sixteen L1 seats went cold in week three. My read is they're not idle — they take the call, run the standard line test, see nothing, and book the two-hour window because that's the only branch the script gives them. The work's happening, it's just happening as a dispatch instead of as a triage decision. But that's my read from the outside. Why do you think they stopped?
Buyer
Because L1 is a BPO working to an average handle time target. You cannot ask an agent on a four-minute AHT to open a second screen and interpret a GPON performance graph. That's not going to happen, and no module you sell me changes it.
Rep
Agreed — and if my answer was 'train the BPO,' I'd be wasting your time. So let me ask the question I actually came with. When we scoped this last year, what did you deliberately cut because it was too big a bite?
Buyer
The upstream classifier. We cut it because none of us believed you could tell an access-side degradation from ours off the same telemetry. Every vendor claims that. And frankly it's the thing that costs us most — half our faults sit on the access network, we don't own them, we just wear the call and the complaint.
Rep
Then that's the conversation. Here's the arithmetic in your numbers, and tell me where it's wrong. Southern ran 6,100 dispatches last quarter, twenty-one percent closed no-fault-found — about 1,280 vans that found nothing. In Northern, just over half of the NFFs carried a signature that sat upstream of the NNI. I'll take half of that to be conservative: 320 avoidable dispatches a quarter. At your loaded one-eighty a dispatch, fifty-seven thousand a quarter. The classifier is seven-four a month for the three Southern regions — twenty-two a quarter. Call it two and a half times, before you count the first-time-fix movement or the complaints you don't get from 'we sent someone and they said it wasn't us.'
Buyer
Your capture rate's generous. Make it a quarter, not a half, and it's still positive but it's not a story I can walk into the exec with. And I meant what I said about engineering — I have no cycles this year.
Rep
Take it at a quarter, I'm happy to run it at your number — it's still net positive and I'd rather you present a conservative one that survives the CFO than mine that doesn't. On cycles: the ask is two hours from whoever owns the northbound feed to point us at the same publisher we already read in Northern, and thirty minutes a week from your NOC lead to set the precision threshold. No integration sprint, no change window, nothing that touches the migration. If it needs more than that before you see a number, I've mis-scoped it.
Buyer
And where does the money come from? Because if you're telling me it comes out of field services, that's Dave's line, not mine, and I'm not walking into that conversation for you.
Rep
You shouldn't have to. Here's what I'd propose instead: give me Southern for one quarter, on your capture assumption. You set the pass mark — my suggestion is upstream faults raised to wholesale before a van moves, plus NFF rate on flagged dispatches. If we miss it we pull it out and there's nothing further to pay. If we hit it, we co-term it to your existing renewal so it's an amendment, not a new purchase, and you and Dave have a number from your own network to split the line with. Does that get you into the room without asking your peer for money on my say-so?
Buyer
That I can take to Dave. Send me the recap with the math at twenty-five percent, and put the engineering ask in hours and names so I can show him it isn't a project.
Objections you will hear
What they say, and what you say back.
| Objection | How to answer it |
|---|---|
| “The capex committee meets quarterly and nothing moves faster than that.” | Then we shouldn't be a capex item. This sits in opex against dispatch cost — if it takes even a quarter of your avoidable truck rolls out, it pays from the field services line, not the network build, and it doesn't compete with the tower program or the node replacements. And realistically, if the committee is eleven weeks out, that's eleven weeks we could spend proving the number in one region, so you walk in with your own dispatch data instead of my slide. |
| “We've already got an assurance platform and event correlation. Why would I add another alarm source?” | You wouldn't, and that's the point — if this spawned a parallel alarm queue your NOC would mute it by week three, and you've watched that happen before. Correlation tells you what already broke. What this adds is a lead indicator on degradation before the customer calls, delivered into the ticket your team already works. And I'd want the precision threshold agreed with your NOC lead up front, in writing: if it fires on noise, we kill it, not defend it. |
| “Half our faults sit on the access network. We don't own the fault, so we can't fix it — we just wear the customer call.” | Right, and that's exactly the truck roll you shouldn't be sending. If you can tell before dispatch that the degradation pattern sits upstream of the NNI, you raise the wholesale fault, you tell the customer the truth on the first call instead of the third, and the technician stays in the van. Fewer no-fault-founds, and fewer of the 'we sent someone and they said it wasn't us' conversations — which is where a serious share of your ombudsman complaints originate. |
| “Our fault and field data is a mess. Half the closure codes are 'other' and the inventory doesn't match what's actually in the ground.” | True at nearly every carrier, and it's the reason the prediction is worth something — the signal comes from performance and alarm telemetry, not from closure codes. Bad inventory affects where we route the dispatch, not whether we spot the degradation. And one of the first things this produces is a list of where the record and the telemetry disagree, which is a free inventory reconciliation you didn't ask for. |
| “We're mid-migration and my engineering team has zero spare cycles this year.” | Then the ask has to fit in the gaps, and I think it does. Two hours from whoever owns the northbound feed — the same publisher we already read in your live region — and thirty minutes a week from your NOC lead to tune the threshold. No change window, no write-back into the ticket flow, nothing that goes near the migration path. If it needs a project team before you see a number, I've designed it wrong and I'd rather find that out today. |
| “Send me something and I'll take it to the exec. We look at this stuff once a year.” | Happy to, but a deck won't survive that room — the CFO will ask what it does to cost to serve per service and nobody in the meeting will have the number. Give me one region's fault and dispatch history for last quarter and I'll come back with how many of those dispatches carried an upstream signature and what they cost you. Then you're presenting your own network's data with your own capture assumption, and I'm not in the room to be argued with. |
| “Honestly, I'm not convinced we're getting full value from what we already bought.” | Then let's spend the call on that and park everything else. Tell me which part isn't landing — is it that the flags aren't reaching the people who make the dispatch decision, or that the NOC doesn't trust the precision? Because if it's the first, that's an onboarding job and I'll bring my CS lead back with a plan and no invoice attached. I'd rather fix the thing you're already paying for than sell you a second one on top of it. |
Questions reps ask about this call
- How do I open an upsell call with a telecom customer without sounding like I'm pitching a new module?
Open with their deployment data, not your roadmap. Something like: "On tickets in Northern where our flag fired before the customer called, no-fault-found closed at nine percent; where it didn't fire, twenty-four. Does that match what your Director of Field Operations sees, or am I reading it optimistically?" Two things happen — you prove you looked before you dialled, and you invite a correction. When a GM Service Assurance corrects your number, they've just told you which metric they defend internally, and that's the metric your business case has to be written in. Never open with "how's everyone finding the platform?" They know you have the usage data in your system, and asking them to narrate it is the fastest way to get demoted from partner to vendor.
- What usage data do I need before running an upsell script for Telecommunications customers?
Six things. Adoption by function — seats vs. weekly actives, split between NOC, L1 triage and field, with the dark corners named out loud. Depth — which screens they work and which they've never opened, plus your exec sponsor's last login. One outcome number sourced from their data, not your benchmark: NFF rate flagged vs. unflagged, truck rolls per 1,000 services deployed vs. control region, or the share of degradations detected before the first customer call. Commercials — value, renewal date, signer, whether the signer is still there, co-term options. Assurance history — any MSD in the last 60 days, any period where your alerts were muted during a fault. And champion status, because telco org charts move after every migration milestone. The test: can you say one thing about their network they'd be mildly surprised you knew?
- The buyer says capex is committed and the committee doesn't meet for another quarter. Is that a real objection or a stall?
It's usually real, and it's usually the wrong budget. Network capex is committed annually against core upgrades, tower builds and node replacements — you will not get in there mid-cycle. The move is to reposition as opex against an existing operational line, almost always field services dispatch cost or contact centre cost. Say it plainly: "This shouldn't be a capex item. If it removes even a quarter of your avoidable truck rolls, it pays from the field services line." Then use the committee gap as a proving window rather than a delay: eleven weeks is enough to run one region and let them walk in with their own dispatch numbers. Do not discount to force it into this cycle — that teaches them the original contract was overpriced and to wait you out on every future expansion.
- How do I handle "we barely use what we've already bought" on a telco account?
Say it before they do, then reframe unused seats as unrouted work rather than wasted spend. "You've got 40 seats, 19 weekly actives, all in the NOC — the L1 desk went cold in week three. If I were you I'd ask why I'd spend another dollar." Then give your read: those agents aren't idle, they're running a line test, seeing nothing and booking the two-hour appointment window, because that's the only branch the script offers. Crucially, be honest about whether the expansion actually fixes that. If the answer is "it doesn't — your L1 is a BPO on a four-minute AHT target and no module changes that," say so. Conceding the point costs you this quarter and buys you the next one. Arguing with it costs you the account.
- What's the right size of ask on a first expansion call in telecom?
One region, one quarter, success criteria written by them on the call, and co-termed to the existing renewal. Not the full footprint, however much your quota wants it. The pilot ask should include a stated exit: "If we miss the pass mark, we pull it out and there's nothing further to pay." Let the buyer choose the metric — upstream faults raised to wholesale before a van moves, NFF rate on flagged dispatches, mean time to detect ahead of first customer contact. Co-terming is the underused unlock: it converts a new purchase into a contract amendment, skips a procurement cycle, and means the CTO or COO signs one renewal instead of negotiating twice.
- When should I park the upsell entirely?
When there's been an MSD or major POI incident in the last 60 days with complaints still open; when your alerts were muted or ignored during a recent fault and nobody has debriefed it; when your champion has just moved in a post-migration reorg; when renewal is inside 60 days with an unresolved value question; when the weekly-active curve is flat or declining and you can't explain why; or when a cost-out program landed this quarter. Say it out loud: "Then this isn't the call — let's use the time on the muted alerts during the POI event, and I'll come back to Southern when that's closed." Pushing an expansion over the top of an open escalation is how a renewable account becomes a churn risk for the sake of an incremental deal.