This week's data pull flagged a small cluster of companies doing the same thing quarter after quarter: clearing consensus with unusual consistency. Not one-off beats, but sequences long enough to suggest something structural beneath the surface.
Using the Financial Modeling Prep Earnings Surprises Bulk API, we screened across the full earnings dataset to isolate repeat outperformers rather than isolated surprises. This article breaks down what that signal is capturing — and how to systematically surface it.
Five Companies With Long Earnings Beat Streaks
Zscaler, Inc. (ZS)
Beat Streak: 32 quarters.
Next quarterly report: June 4 — EPS: $0.99; Revenue: $835.77M (consensus).
A 32-quarter beat streak places Zscaler in a category where the signal is less about episodic upside and more about systematic underestimation. At that length, the pattern reflects a recurring gap between how quickly analysts update models and how consistently the company converts demand into reported results. In cybersecurity, where revenue visibility is often tied to multi-year contracts and deferred revenue, this kind of streak frequently aligns with strong forward booking visibility and disciplined cost scaling. The persistence suggests internal forecasting precision is outpacing external expectations rather than simply benefiting from favorable quarters.
What stands out is how the streak has held through multiple demand environments — from elevated enterprise security spending cycles to more scrutinized IT budgets. That continuity points to execution stability rather than reliance on macro tailwinds. To contextualize this further, deferred revenue trends and remaining performance obligations (RPO) from the earnings filings would help quantify how much future revenue is already embedded in contracts. If the beat pattern is tied to backlog conversion rather than new sales acceleration, the signal shifts from growth momentum to revenue predictability — a different, but equally important, form of consistency to monitor.
Celestica Inc. (CLS)
Beat Streak: 26 quarters.
Next quarterly report: April 23 — EPS: $2.07; Revenue: $4.05B (consensus).
Celestica's 26-quarter streak emerges from a segment of the market where expectations tend to be cyclical and tightly linked to end-market demand variability. Electronics manufacturing services typically operate on thinner margins and are sensitive to supply chain dynamics, making long beat sequences less common. That makes the signal here more indicative of operational control — particularly around cost management, customer mix, and capacity utilization — rather than simply revenue upside.
The durability of the streak suggests that Celestica has been able to navigate shifts in demand across industrial, communications, and enterprise segments without significant earnings volatility relative to estimates. In these environments, small improvements in margin discipline can consistently translate into earnings outperformance. Segment-level margin data and backlog disclosures from earnings reports would help clarify whether the consistency is being driven by mix shift (higher-margin programs) or structural efficiency gains. Either way, the repeatability indicates that consensus models may be underweighting execution stability in a business typically viewed as cyclical.
Palantir Technologies Inc. (PLTR)
Beat Streak: 10 quarters.
Next quarterly report: May 4 — EPS: $0.29; Revenue: $1.53B (consensus).
Palantir's 10-quarter beat streak coincides with its transition into sustained GAAP profitability, marking a shift in how its financial profile is interpreted. Earlier periods were often characterized by variability tied to contract timing and stock-based compensation. The current sequence, however, reflects a more consistent relationship between revenue growth, operating leverage, and reported earnings. That shift is significant because it reduces the range of outcomes analysts need to model, yet the streak indicates estimates are still being cleared with regularity.
The signal here appears tied to the scaling of its commercial segment alongside steady government demand, creating a dual-engine structure that smooths revenue realization. Monitoring segment revenue splits and contribution margins in the income statement would provide clarity on whether one segment is driving the majority of upside or if both are contributing evenly. Additionally, tracking customer count growth and average contract value could help determine whether the consistency stems from broader adoption or deeper expansion within existing clients. The pattern suggests a maturing revenue model where predictability is improving faster than consensus adjustments.
The Allstate Corporation (ALL)
Beat Streak: 10 quarters.
Next quarterly report: April 29 — EPS: $7.31; Revenue: $17.3B (consensus).
Allstate's streak operates within a fundamentally different framework than the other names on this list, as insurance earnings are heavily influenced by underwriting discipline, pricing cycles, and claims variability. A 10-quarter sequence of beats in this context often reflects successful rate adjustments and risk selection rather than top-line expansion alone. Given the volatility inherent in catastrophe losses and claims inflation, sustained outperformance relative to estimates suggests that pricing actions have been both timely and sufficient to offset cost pressures.
The consistency is particularly notable given the broader industry environment, where insurers have been recalibrating premiums in response to elevated loss trends. Combined ratio data and underwriting margins from the income statement would be central to understanding how much of the beat streak is driven by core insurance operations versus investment income. If underwriting performance is the primary contributor, the signal points to disciplined execution in a complex risk environment. The pattern highlights how earnings predictability in insurance is less about stability in inputs and more about responsiveness in pricing and portfolio management.
Williams-Sonoma, Inc. (WSM)
Beat Streak: 7 quarters.
Next quarterly report: May 28 — EPS: $1.81; Revenue: $1.79B (consensus).
Williams-Sonoma's seven-quarter streak stands out within consumer discretionary, where demand is typically more elastic and sensitive to shifts in spending behavior. Sustained earnings beats in this sector often reflect a combination of pricing power, inventory discipline, and brand positioning. In this case, the consistency suggests that the company has maintained margin control even as broader retail conditions have fluctuated, indicating that operational execution is compensating for variability in demand.
The signal becomes clearer when viewed through the lens of gross margin and inventory turnover. Retailers that consistently outperform expectations in mixed demand environments are often managing promotions and supply chains more efficiently than peers. Income statement data on gross margins, alongside balance sheet indicators such as inventory levels, would help determine whether the streak is being driven by cost control, product mix, or reduced discounting. The repeatability suggests that consensus estimates may be underestimating the company's ability to preserve margins rather than overestimating revenue growth — a subtle but important distinction in interpreting the signal.
Interpreting What Repeatable Beats Are Actually Telling Us
Across these five names, the common thread is not sector, size, or business model — it is the persistence of expectation gaps. When earnings beats extend into double-digit quarters, the signal shifts from “positive surprise” to something more structural: a recurring mismatch between how companies internally forecast their performance and how the market models it externally.
What emerges is less about growth in isolation and more about forecasting friction. In each case, analysts appear to be adjusting estimates incrementally, while underlying business performance compounds more steadily. That dynamic can stem from different sources — contract visibility in software, margin discipline in manufacturing, pricing power in insurance, or inventory control in retail — but the observable outcome is the same: reported results continue to clear a consensus that is consistently catching up rather than leading.
Looking at the pattern through a broader dataset lens reinforces that point. Earnings surprise data alone identifies the streak, but it doesn't explain its durability. That requires layering in additional context. For example, comparing earnings beats against operating margin trends from FMP's Income Statement API can help determine whether the consistency is driven by structural margin expansion or simply conservative estimates. In parallel, aligning those results with analyst target revisions from FMP's Price Target Summary endpoint shows whether consensus is recalibrating meaningfully — or remaining anchored even as companies repeatedly outperform.
There is also a timing dimension embedded in these streaks. When earnings surprises are mapped alongside historical estimate revisions and subsequent reporting periods, a pattern often emerges: estimates tend to adjust with a lag, particularly when underlying drivers are less visible or harder to model. Pulling multi-period earnings histories through the Earnings Report API and pairing them with forward-looking estimates allows that lag to be quantified rather than inferred. In effect, the dataset reveals whether beats are narrowing over time (suggesting convergence) or persisting (suggesting ongoing misalignment).
Taken together, repeatable earnings beats are less a signal about any single quarter and more a reflection of how efficiently information is incorporated into expectations. When the same companies continue to outperform across multiple reporting cycles — even as their industries, cost structures, and demand environments differ — the takeaway is not that the businesses are uniformly “stronger,” but that the modeling frameworks around them are slower to fully absorb what those businesses are already demonstrating in their reported numbers.
Building a Repeatability Screen with FMP Data
When the goal is to find companies that regularly outperform expectations, the process needs to start with the full dataset rather than a pre-selected watchlist. Beginning with a narrow list of familiar names introduces bias before the analysis even begins. A more reliable approach is to pull the entire universe of reported earnings outcomes and then allow the data itself to reveal which companies repeatedly exceed estimates.
That's where the FMP Earnings Surprises Bulk API becomes useful. It provides a standardized record of quarterly EPS results compared with analyst estimates across a wide range of equities, making it possible to identify patterns in earnings surprises at scale.
As with any automated workflow, the only prerequisite is confirming that your API key is active before making requests.
1. Pull Bulk Earnings Surprises
Begin by hitting the Earnings Surprises Bulk API, which aggregates every quarterly EPS surprise — positive or negative — for the year you specify:
https://financialmodelingprep.com/stable/earnings-surprises-bulk?year=2025&apikey=YOUR_API_KEY
Sample Response:
[
{
"symbol": "AMKYF",
"date": "2025-07-09",
"epsActual": 0.3631,
"epsEstimated": 0.3615,
"lastUpdated": "2025-07-09"
}
]
From here, the first cut is mechanical: isolate the entries where epsActual > epsEstimated. That gives you the universe of names that beat expectations at least once during the period — essentially a raw pool before you evaluate whether any of them can deliver that result consistently.
2. Retrieve Company-Level Details
With that universe in hand, the analysis moves from identifying events to evaluating consistency. For each ticker that cleared the first filter, pull its complete quarterly earnings history using the Earnings Report API:
https://financialmodelingprep.com/stable/earnings?symbol=AAPL&apikey=YOUR_API_KEY
Looking at the complete sequence of reported quarters makes it possible to evaluate frequency and clustering. This is where judgment enters the workflow. Some analysts require three or more consecutive beats to qualify as a streak; others impose minimum surprise thresholds or remove near-zero deviations. The parameters can be adjusted, but the intent stays the same: separate sustained execution from statistical noise.
By this point, the screen has moved beyond identifying isolated surprises. What began as a broad event scan turns into a structured assessment of earnings reliability, highlighting companies where internal forecasting and operational control have proven more consistent than the market's expectations over time.
Broadening the Universe as Coverage Scales
Scaling a repeatability screen works best when the dataset expands in stages. The goal isn't simply to analyze more companies—it's to verify that the screening logic continues to hold as the environment becomes less predictable. A methodology designed to detect persistent earnings beats should first demonstrate stability where information is most complete, before being applied to parts of the market where analyst coverage is thinner and estimates carry wider dispersion.
The logical starting point is the Free plan, where the dataset is largely composed of widely followed large-cap companies such as AAPL, GOOGL, and JPM. These names sit in the most transparent part of the market: analyst coverage is dense, consensus estimates cluster tightly, and earnings expectations tend to be well modeled. If a streak-based screen produces coherent results under those conditions, it suggests the framework itself is sound rather than reacting to data gaps or irregular coverage.
From there, moving into the Starter plan broadens the U.S. universe to include smaller and more specialized companies. In that environment, fewer analysts typically follow each ticker, which means consensus estimates can vary more widely from quarter to quarter. That variability effectively stress-tests the screen. If the same filtering rules still surface repeatable earnings beats, the signal is less likely to be an artifact of tightly modeled large-cap forecasts and more likely to reflect genuine operational consistency.
The next layer comes from geographic expansion through the Premium plan, which introduces additional markets such as the U.K. and Canada. The mechanics of the screen remain unchanged, but the analytical context becomes more complex. Differences in accounting conventions, industry composition, and reporting practices introduce additional variability. Applying identical screening criteria across these regions ensures the methodology is evaluating the same concept of earnings repeatability, regardless of where the company reports.
Viewed as a whole, the staged expansion serves as a practical validation framework. Begin where coverage is deepest, introduce variability through smaller-cap companies, and then extend the analysis internationally. When the signal continues to appear at each stage without requiring constant recalibration, the likelihood increases that the screen is identifying a genuine pattern rather than a quirk in a particular dataset.
When a Screening Process Becomes Research Infrastructure
When a screening workflow consistently holds up across several earnings cycles, its relevance naturally extends beyond the analyst who built it. What begins as a desk-level process for identifying repeatable earnings beats often evolves into something broader: a framework the entire research organization can use to evaluate earnings consistency. At that point, the question shifts from “Does this help my coverage universe?” to “Should this become part of how the firm measures operational reliability across sectors?”
In practice, those transitions are rarely initiated from the top down. They tend to emerge from analysts who have worked closest with the data—refining filters, resolving edge cases, and pressure-testing assumptions across multiple reporting seasons. As those definitions stabilize, they offer a path away from fragmented workflows: separate spreadsheets, slightly different sector methodologies, and competing interpretations of the same concept. A shared screening framework replaces that fragmentation with a common analytical baseline that teams can apply consistently across coverage groups.
Standardizing the workflow changes the mechanics of internal research. Shared dashboards replace isolated models maintained by individual desks. Adjustments to thresholds or definitions become visible and reviewable rather than embedded in personal files. Auditability improves because the inputs, filters, and calculations are explicit rather than implied. Governance follows naturally from that transparency. Instead of spending time reconciling why two teams arrived at different numbers, analysts can focus on interpreting what the data actually signals.
Once a workflow reaches that stage, scaling it becomes less about convenience and more about maintaining methodological integrity. Centralized infrastructure—such as the Enterprise plan—allows a process already validated at the desk level to operate across teams with consistent data access, version control, and shared visibility. The objective is not to change the analysis itself, but to ensure that as adoption grows, the logic behind the screen remains consistent across the entire research organization.
Keeping Earnings Consistency in Motion
Consistency at this level is less about isolated performance and more about how reliably a company converts its operating model into outcomes that exceed expectations. Tracking that signal at scale — using datasets like the Earnings Surprises Bulk API — keeps the focus on what is actually being delivered, not just what is being projected. Over time, the edge comes from recognizing when that gap persists rather than closes.
Want more? Explore our earlier article: Weekly Signals Desk | Five Notable Price-Target Gaps via the FMP API (March 9-13)
Disclosure: Signals Desk content is provided for informational and analytical purposes only and does not constitute investment advice or trade recommendations. The analysis reflects interpretation of market data and publicly disclosed or third-party information, including data accessed via Financial Modeling Prep APIs, at the time of publication. Signals discussed are probabilistic, can be wrong, and may change as market conditions and consensus data evolve. This content should be considered alongside broader research, individual objectives, and risk assessment.
