Headline
Client identity withheld under NDA. All figures sourced from live campaign data. Running period: February-March 2026.
For iGaming operators managing a maturing player database in a regulated market, player reactivation is where the most recoverable revenue sits. This case documents how a leading Portuguese casino operator deployed Bswan's AI-driven retention platform across two sequential campaigns against the same AWOL segment – and what happened to performance when deposit-based segmentation and extended messaging sequences were introduced between runs.
Want to run your first reactivation campaign?
Let's map your AWOL segment and project the recoverable NGR.
Get a quote
Results at a Glance
Combined 2-month NGR: $61,676
Total spend (incl. one-time setup): ~$13,094
Cost per activated player: $0.293
The Challenge. Revenue Sitting in an Overlooked Segment
Portugal's regulated online casino market is one of the most competitive in Southern Europe. For operators at this scale, a database of 23,000 players in a single AWOL segment represents material recoverable revenue – the kind that standard CRM channels consistently under-deliver against.
This operator had the offers, the budget, and the retention team. What their existing stack delivered was reach without resonance: messages landing in inboxes players had stopped checking, bonus offers presented without reference to individual player deposit behaviour, and a channel mix that required high volume to compensate for low engagement rates.
Outbound player reactivation via AI voice changes the dynamic of that equation. A call reaches a player in real time. A conversation — even a 30-second AI-driven one – carries a level of personalisation that batch email sequences structurally cannot replicate. The operator needed a way to activate that potential at scale, with offer logic matched to individual player lifetime value, and with infrastructure that ran without adding headcount to their team.
The Bswan Approach: AI Voice + SMS Built Around Player Value
Bswan deployed two sequential AI-driven retention campaigns against the same AWOL player pool, each structured as a self-contained project: a configured AI voice agent, a tiered SMS follow-up sequence, and a validated contact pipeline – operating on a usage-based model where the operator paid per valid player contact processed.
Before either campaign reached a single player, the full 23,000-record base went through Bswan's phone number validation layer. Of those, 19,516 came back as valid and contactable. The operator's cost applied only to that validated pool – every unreachable or duplicate record was excluded from billing before outreach began. This validation step is standard across all Bswan campaigns and directly shapes the unit economics of casino reactivation campaigns at scale.
“We run retention across multiple funnels and campaign types. AI voice is not our only tool, but for AWOL segments, the numbers it produces are consistently stronger than anything else we run in that part of the funnel.”
Campaign 1. Establishing the Reactivation Baseline
Running period: 13.02.26 – 25.02.26
Funnel: AI voice intro call → SMS follow-up
The AI voice agent initiated a personalised call to each of the 19,509 reached players. Where pickup was confirmed, the agent delivered a scripted intro with a first-deposit bonus offer calibrated to the Portuguese-speaking regulated market. The SMS follow-up sequence ran in parallel – reaching players who engaged with the call but required a second touchpoint before acting, as well as those the call did not reach live.
Voice as a player reactivation channel carries a structural advantage over batch messaging: a call reaches a player in real time, at the individual level, with a contextual offer. That combination produces engagement rates that email-only iGaming reactivation programmes consistently fall short of, particularly on AWOL cohorts beyond the 90-day mark.
The combined funnel produced:
A 2.47% casino reactivation rate on a 120+ day AWOL cohort – a segment most operators have already discounted in their revenue forecasts – represents a recoverable return that the operator's existing channel mix had failed to unlock.
Campaign 2. Segmentation, Extended Sequences, and a Doubled Conversion Rate
Running period: 25.02.26 – 30.03.26
Funnel: AI voice intro call → SMS follow-up → SMS follow-up 2
Campaign 2 ran against the same 23,000-player base with two structural changes applied between runs.
Structural change one: deposit-tier segmentation. The player base was divided into value tiers based on historical player deposit behaviour. Each tier received a distinct offer sized to match that player's demonstrated deposit comfort zone. A player with €8,000 in lifetime deposits received a materially different offer than a player with €150 in bonus size, minimum deposit threshold, and game title. The principle was straightforward: bonus segmentation by player value eliminates the conversion drag of presenting a high-value player with a low-value offer, and vice versa.
To illustrate the range: the entry segment (under €200 lifetime deposits) received a €10 deposit prompt with 50 Free Spins, while the top segment (€25,000+) received a €50 deposit prompt with 250 Free Spins on a premium title. Eight distinct offer variants ran simultaneously across the same AI voice infrastructure, with personalisation variables handled inside the existing campaign structure – no additional setup required.
Structural change two: extended SMS sequence. A second SMS follow-up was added to the funnel, creating an additional decision window for players who engaged with the first touchpoint but did not act immediately. The 33,205 SMS messages sent in Campaign 2 – versus 19,522 in Campaign 1 – reflect both the larger outreach volume and the extended sequence depth.
Campaign 2 results:
7,213 full conversations completed
2,051 unique links clicked – a 133% increase over Campaign 1
999 activations at 5.38% conversion
$42,328 NGR over 60 days
The conversion rate more than doubled. Link click volume more than doubled. NGR from Campaign 2 alone was 2.2x the NGR from Campaign 1. The same player base, the same AI voice infrastructure, the same usage-based cost structure – the performance difference came entirely from offer-to-player matching and extended sequence depth.
Why the Results Compounded
Three structural factors drove the performance gap between Campaign 1 and Campaign 2.
AI voice as a reactivation channel operates differently from batch messaging because it works in real time at the individual level. When players answered the phone in Campaign 1, each received a call that addressed them by name, in Portuguese, with a contextual offer. That combination – real-time channel, personal address, relevant offer produces player engagement rates that email-only iGaming reactivation programmes consistently fall short of, particularly on AWOL cohorts beyond the 90-day mark.
Bonus segmentation by deposit tier was the single biggest conversion lever between campaigns. Offer-to-player fit matters more than offer size. A uniform bonus distributed across a heterogeneous player base optimises for neither the low-value nor the high-value segment. Segmenting by demonstrated player deposit behaviour and calibrating the offer accordingly is what moved the conversion rate from 2.47% to 5.38%.
Sequential touchpoint depth created the additional decision window that lifted link click volume by 133%. Players who received the first SMS and engaged but did not convert were still within a recoverable consideration window. The second follow-up reached them. This is a structural property of multi-touch sequences – adding depth to the funnel captures a portion of engaged players that a single-touch campaign leaves behind.
The compounding effect between campaigns came from applied learning. The operator used Campaign 1 conversion data to build the segmentation logic for Campaign 2. Bswan carried the same validated base, the same AI voice framework, and the same delivery infrastructure forward. The operator layered refined offer logic on top of a proven system, which is precisely how iGaming player lifetime value compounds across sequential reactivation programmes.
What This Operator Actually Paid: The Usage-Based Model
Bswan's usage-based pricing for this campaign operated across two cost layers.
Setup (one-time): $1,950. The Essential package covered AI voice script development, SMS copy preparation, campaign structure and flow configuration, platform account creation, and player upload setup. Every AI voice agent script was reviewed and approved by the operator's team before a single call went live. Client sign-off on the script is a standard requirement – the operator controls what the AI says before any outreach begins.
Processing (per valid player contacted). The operator paid a per-contact processing fee for each valid phone number reached across both campaigns. This fee covers telephony, AI compute, SMS delivery, and all tracking infrastructure. The blended processing spend – $5,713.50 for Campaign 1 and $5,430.70 for Campaign 2 – represents the full cost of reaching 19,509 and 18,535 players, respectively, across voice and SMS combined.
The economics of that spend against revenue: $13,094 total investment generated $61,676 in NGR across two campaigns run over 60 days combined.
Under the usage-based model, campaign spend scales with the size of the contacted base. Operators with larger AWOL segments spend more in absolute terms; the per-player unit cost remains consistent. What is included by default in the processing fee: personalisation variables across the player base, multiple offer variants within the existing script structure, bonus segmentation logic across value tiers, and the validated phone number pipeline. The operator running Campaign 2's eight-tier offer segmentation across 18,535 players incurred no add-on fees for that segmentation depth – it operated within the existing campaign configuration.
The setup fee carries a full money-back guarantee within the first 30 days from campaign launch if the campaign does not meet performance expectations.
See what AI-driven player reactivation looks like for your market.
What the Retention Team Operated
The operator's retention team defined the offer logic, reviewed the AI voice scripts, and approved the campaign structure before launch. Bswan's platform handled phone number validation, call routing, voicemail delivery, SMS sequencing, link tracking, and conversion attribution.
The resource investment from the operator's side across both campaigns was the time to review scripts before launch and to analyse results between runs. The infrastructure ran inside Bswan's system. The team's attention stayed on offer strategy and performance decisions rather than campaign execution.
Player data upload was via Bswan's platform UI. API integration is available for operators requiring automated real-time syncing. The platform account was created as part of the setup package the operator arrived at a configured environment ready for the first campaign run.
For operators concerned about regulatory exposure in the PT market: every script version receives explicit client sign-off before deployment. There is a documented approval trail for every AI voice agent version that goes live. The operator retains full visibility and control over what the AI communicates to their players.
The Compounding Case for Sequential Campaigns
The most commercially significant data point in this case is the relationship between Campaign 1 and Campaign 2.
Campaign 1 confirmed that AI voice reactivation worked for this operator's AWOL segment in the Portuguese market at 2.47% conversion. Campaign 2 applied two refinements to that confirmed foundation – deposit-tier segmentation and an extended SMS sequence, and the conversion rate moved to 5.38%. NGR moved from $19,348 to $42,328.
The same player base. The same AI voice infrastructure. The same usage-based cost. The performance difference was entirely structural, and both structural changes are available to any operator running a second campaign on an established Bswan setup.
For iGaming operators with a maturing player database, the compounding economics of sequential AI-driven retention campaigns become the structural argument for treating player reactivation as an ongoing programme rather than a one-off campaign. Each run produces conversion data. That data informs offer construction. Better offer construction produces higher conversion rates. Higher conversion rates improve the return on the per-player processing fee.
This operator generated $61,676 in NGR across two campaigns. The trajectory from here – with an approved AI voice agent, a validated base, and an established segmentation model already in place – is a programme that performs better with each iteration, on the same usage-based cost structure.
