#104 - TSFS: Year Two - what you missed

Somehow, the improbable thing happened. Another year has passed.

Another year of fighting fraud, helping clients grow, and writing. A lot of writing.

I remember when I was afraid I'd run out of ideas by issue 15 of TSFS. But it’s been over 100 issues now, and I know I can keep going forever.

Not to mention, TSFS is now a podcast!

Yup, this year has been unbelievable as it’s been a stretch.

For those of you who’ve been there - thank you!

I cannot stress enough how much your feedback and insights fuel my writing.

And for those of you who just joined the ride - welcome!

I’m so happy to see you here.

To celebrate this 2-year anniversary, and as I did last year, I wanted to share a quick reference guide to the last 52 issues of TSFS.

If you joined mid-ride, if you want a refresher, or if you want to share it with your closest loved ones, this is it.

Side note: I highly recommend bookmarking this page on my website so you can always find this list quickly.

So without further ado, here’s everything from “TSFS: Year Two”:

Strategic Frameworks & Defense Architecture

Issue 56: Fraudsters limit exposure too. Here's how to exploit it. How account limits force fraudsters to scale their attacks, and how to turn limit-testing behavior into detection signals.

Issue 57: Stop trying to predict fraud. Do this instead. Why “proactive” fraud prediction is a myth, and a one-hour-a-week framework (Platform, People, Process) for building real readiness.

Issue 61: Real-time fraud defense has a fatal flaw Three ways batch analysis catches what real-time systems can't: whole fraud rings, siloed data, and slow-but-predictive signals.

Issue 63: Why sophisticated fraud defenses fail more often Why over-engineered rules and “fraudsters can bypass it” objections backfire, and how to keep your defenses simple and boring.

Issue 88: Duct tape beats a "perfect" vendor Why risk leaders need to switch to “MacGyver mode” when a crisis hits, patching with what they have instead of saying “I told you so”.

Issue 99: What is fraud strategy for? Why fraud strategy is about building capacity to react, and how skills, systems, processes, and data can relieve an overloaded team without adding headcount.

Issue 102: Fraud targets are broken. Ask for a budget. How to negotiate a decline budget with the business using a menu of risk appetites priced in revenue and losses, then manage it like a portfolio.

Team & Organizational Design

Issue 54: 99% of fintechs have this fraud blind spot Why siloed identity and payment fraud teams let fraud slip through the cracks, and three low-drama ways to connect them.

Issue 76: Cut costs with AI - without losing your team or your CFO A framework for making the AI cost-cutting case: map your costs first, pitch decoupling OPEX from growth, and give your team a real plan.

Issue 84: The AI hangover no one's preparing for Why agentic AI running at machine speed turns governance into a management problem, and why your Fraud Analytics function needs to grow, not shrink.

Issue 94: How fraudsters (accidentally) find your gaps Why attacks succeed in the gaps between teams rather than inside them, and how naming an owner and a metric closes those gaps.

False Positives & Approval Rates

Issue 53: Your approval rate metric is lying to you Why raw approval rates mislead executives, and how a Fraud-Adjusted Approval Rate (FAAR) reveals your team's real performance.

Issue 67: Your $50 decline just cost you $5,050 How to calculate the lifetime-value cost of a false positive, and how to protect the two customer segments that punish friction the most.

Issue 80: How adding friction increased card spend by 11% Data from Novo showing how well-placed “tactical friction” builds customer trust and increases spend instead of driving churn.

Rules & Detection Logic

Issue 66: Are you using these 3 rule types? The difference between strategy, behavioral, and trend rules, and how to own, monitor, and measure each one differently.

Issue 79: The ML score cheat code almost no one uses in rules Two ways to wire ML scores into behavioral and trend rules, including a low-score exclusion trick that quietly cuts false positives.

Issue 85: Stop letting fraudsters write your rules How to use your existing rules as a map for proactive expansion, across new use cases and new resolution modes, instead of waiting for the next attack.

Issue 95: The 3 myths behind rules vs. AI The three myths that make teams apologize for being rules-based, and why rules and AI work best together.

Machine Learning & Models

Issue 72: The self-learning fraud model is a lie Why “self-learning” vendor claims can't survive label lag, and the retraining question to ask instead.

Issue 78: Are you actually using your ML score? How to set ML score thresholds by defining your goals, segmenting your population, and letting AI run the analysis.

Issue 86: Would you hire a fraud analyst who knows everything? Why granular solutions beat one-size-fits-all, from specialized AI agents to approval rules to models trained on your own data.

Issue 91: Have foundation models dethroned ML? A close read of the fraud foundation model claims from Stripe, Mastercard, Revolut, and Coinbase, and why foundation models will enrich ML rather than replace it.

Issue 97: Did foundation models just end fraud experts? What Sardine's issuing foundation model reveals about how these models actually work, and whether they make fraud domain experts obsolete.

Testing, Monitoring & Labeling

Issue 70: You might be testing your solutions blindly When to use backtesting, Champion-Challenger, and shadow mode, and how to layer them before a new solution goes live.

Issue 71: The fraud you blocked will break your next model Why unlabeled blocked fraud quietly weakens your next model, and practical ways to label it without full automation.

Issue 73: Your fraud system is dying Why every fraud solution decays over time, and how to catch degradation by measuring against go-live performance.

Issue 77: AI will fix fraud - just not how you think Why AI's biggest impact on fraud is continuous labeling that speeds up the reaction cycle, and where it must stay supervised.

Issue 87: Your vendor isn't watching. Your team isn't either. Why neither vendors, leadership, nor human reviewers will catch AI agent drift, and the three metrics someone on your team should own instead.

Fraud Investigations & Signals

Issue 55: Is it 1st or 3rd party fraud? (and why it matters) How to tell first-party from third-party fraud, and why using the wrong tools for each costs you millions.

Issue 81: Stop investigating fraud. Start disproving it. Why open-ended investigation protocols create rabbit holes, and how a “decide first, then disprove” model gives every case a finish line.

Issue 82: Confidence sabotages your fraud investigation A step-by-step ATO case walkthrough of the disproval method, and the question that tells you when to stop investigating.

Issue 83: Good customers are messy. Fraudsters are not. Why a missing digital past, flawless form inputs, and too-efficient sessions are some of the strongest signs you're facing a fraudster.

Issue 93: The two things fraudsters can't hide The two fingerprints every fraud ring leaves, intent and infrastructure, and how micro-segment monitoring surfaces them early.

Issue 101: The ring you just blocked is innocent How to read the links between accounts by uniqueness and by identity vs. user, so you don't block a whole ring of good customers.

AI Adoption & Strategy

Issue 59: The AI mistake fintech leaders keep making Why using AI for efficiency caps your ROI, and how to use it to unlock new capabilities instead.

Issue 60: The question I stopped asking in 2025 Why “Will AI replace me?” is the wrong question for fraud practitioners, and the one that actually protects your career.

Issue 90: You are not ready for AI Why AI needs better data hygiene than your analysts do, and three preparation steps most teams skip (with Shachar Meir).

Issue 96: The 4 steps to roll out AI The difference between AI activation and AI enablement, and a four-point bar for rolling AI out as a program rather than a project (with Brian Davies).

Issue 100: Fraud teams are the worst AI builders Why building your own agents is a trap without engineering scaffolding, and four questions to answer before you build.

Vendor Selection & Management

Issue 62: What you lose when you choose flexible fraud integrations Why skipping status callbacks and label feedback during integration caps your vendor's performance, and how to stop “phase two” from becoming “never.”

Issue 65: 5-min guide to entity resolution tools The questions that separate real-time graph entity resolution from pretty investigation visuals: real-time limits, templates, hop depth, fuzzy linking, and common assets.

Issue 69: You might already be too late A 12-month vendor renewal framework so you never get locked into a vendor you're unhappy with for another contract cycle.

Issue 89: Give me 5 minutes, I'll tell you if your vendor's worth it. A five-minute, back-of-the-envelope method for sanity-checking vendor ROI before a renewal decision.

Emerging Threats & Agentic Commerce

Issue 64: 3 Agentic Commerce questions nobody's asking Three overlooked fraud questions about agentic commerce: abandoned step-up flows, traffic shifting to mobile web, and fraud stacks breaking under the strain.

Issue 74: Fraud is down. Losses are up. Here's why. What the MRC 2026 report's “fewer incidents, higher losses” paradox suggests about fraud migrating to FIs and rising returns abuse.

Issue 75: KYA is overhyped Why Know Your Agent only solves agent spoofing, and the harder agentic commerce problems it leaves untouched.

Issue 92: AI fraud is here. So where's the loss? Why AI-powered fraud is hitting consumers' phones instead of institutions' dashboards, and why that may not last.

Stories & Industry Insights

Issue 58: Why am I Joining Sardine Why I joined Sardine after two and a half years of solo consulting, and what it means for Native[risk] and this newsletter.

Issue 68: Fraud is good for business What LexisNexis' True Cost of Fraud study reveals about how fraud pressure pushes teams to build better customer experiences.

Issue 98: Who is The Saturday Fraud Strategist? The story behind the newsletter's name and its Saturday slot, and who The Saturday Fraud Strategist really is.

In the meantime, that’s all for this week.

See you next Saturday.

P.S. If you feel like you're running out of time and need some expert advice with getting your fraud strategy on track, here's how I can help you:

Free Discovery Call - Unsure where to start or have a specific need? Schedule a 15-min call with me to assess if and how I can be of value.
​Schedule a Discovery Call Now »​

Consultation Call - Need expert advice on fraud? Meet with me for a 1-hour consultation call to gain the clarity you need. Guaranteed.
​Book a Consultation Call Now »​

Fraud Strategy Action Plan - Is your Fintech struggling with balancing fraud prevention and growth? Are you thinking about adding new fraud vendors or even offering your own fraud product? Sign up for this 2-week program to get your tailored, high-ROI fraud strategy action plan so that you know exactly what to do next. ​
Sign-up Now »

Enjoyed this and want to read more? Sign up to my newsletter to get fresh, practical insights weekly!

<