Why Google Switched from FID to INP

TL;DR

Google replaced First Input Delay (FID) with Interaction to Next Paint (INP) as a Core Web Vital on March 12, 2024. FID measured only the browser's delay before processing the first user interaction. INP measures the total time from any user interaction — click, tap, or key press — to the first visual response on screen, across all interactions during a session, reporting the worst one. INP is harder to pass: over 30% of mobile sites still fail the threshold. More importantly in 2026, poor INP scores affect more than organic rankings. Page responsiveness directly influences crawlability by AI crawlers — including GPTBot and PerplexityBot — and the user experience signals that predict AI citation eligibility. This guide explains FID vs INP in full, and connects INP optimization to the AI search visibility layer that Dageno AI monitors.

What FID and INP Have in Common

Both metrics measure website responsiveness — how a page reacts to user actions. Both are based on real user interactions: neither FID nor INP is recorded if a user opens a page and reads without clicking, tapping, or pressing a key. You cannot reliably measure either in pure lab environments by reloading a page; you need to simulate real interactions to generate scores.

This distinguishes both from LCP (Largest Contentful Paint) and CLS (Cumulative Layout Shift), which are captured automatically on page load regardless of user behavior.

The Key Differences: Why INP Replaced FID

1. First Interaction vs. All Interactions

FID measures only the first user interaction — the delay before the browser begins processing the very first click, tap, or key press after page load.

INP collects data on every interaction throughout the entire session, then reports the worst single interaction — the one that took the longest to produce a visible response. Google applies one exception: if a page has 50+ interactions in a session, the single highest outlier is ignored, preventing rare catastrophic events from distorting the score. For most pages with fewer than 50 interactions per session, the longest interaction becomes the INP score.

Practical example of how INP updates during a session:

• First interaction: 104ms → INP = 104ms
• Second interaction: 472ms → INP updates to 472ms
• Third–tenth interactions: 16ms–56ms → INP remains 472ms
• The 472ms interaction was the worst; it defines the session INP

2. What Is Actually Being Measured

FID measures only input delay — the time from a user action to when the browser begins processing it. The browser's processing time and the visual rendering of the response are not included in FID.

INP measures the complete interaction duration: input delay + processing time + presentation delay — the time until the first visual frame change appears on screen after the user's action. INP reports the total time the user waits to see that something happened.

This makes INP a fundamentally more complete measure of interaction quality. A page that quickly starts processing an interaction but takes 800ms to visually update could pass FID (fast input delay) while failing INP (slow total interaction duration).

3. Pass Rate and Difficulty

The FID transition was partly driven by the fact that most websites had already achieved good FID scores — the metric had become insufficiently discriminating. Over 90% of websites were passing FID on both mobile and desktop, making it a poor benchmark for differentiating interaction quality across the web.

INP is significantly harder to pass. More than 30% of mobile sites still fail the INP threshold, providing a more meaningful differentiation signal and a genuine optimization target for the majority of the industry.

4. User-Centricity

FID's first-interaction-only scope means it could produce a misleading picture of user experience. A page that responds instantly to the first click but becomes progressively more unresponsive as the user interacts deeper — perhaps because of JavaScript execution accumulating on a heavy interactive page — would show an excellent FID while delivering a poor actual experience.

INP, by measuring all interactions and reporting the worst, reflects what users actually experience throughout their time on the page — making it a more accurate proxy for user-perceived responsiveness.

The INP Thresholds

A good INP score requires that every interaction across a session completes within 200ms from user action to visual response. This is significantly more demanding than FID, where only the first interaction's input delay needed to be fast.

Why This Matters Beyond Traditional SEO: INP and AI Crawler Access

In 2026, Core Web Vitals optimization has implications beyond organic rankings that were not present when FID was introduced.

According to Cloudflare's May 2025 crawler analysis, GPTBot has grown to 7.7% of all crawler market share (up from 2.2% in May 2024, +305%) and PerplexityBot has grown 157,490% from minimal presence. Together with ClaudeBot and other AI crawlers, AI bot traffic now represents a significant and growing fraction of total crawler activity.

AI crawlers do not execute JavaScript. GPTBot, ClaudeBot, and PerplexityBot all consume static HTML only. Heavy JavaScript payloads that produce poor INP scores — because they block the main thread, delay event processing, or create long tasks that prevent timely visual updates — are also the payloads most likely to make your content invisible to AI crawlers. The same JavaScript optimization that improves INP for real users also reduces the rendering burden that causes AI crawlers to receive empty or malformed content.

This creates a direct alignment: fixing the JavaScript issues that cause poor INP scores also improves the AI crawler accessibility that determines whether AI platforms can index and subsequently cite your content.

Tools for Measuring INP

• Google Search Console — Core Web Vitals report with field data from real Chrome users
• Google PageSpeed Insights — combines field data (CrUX) with lab data for individual URL analysis
• Chrome DevTools — Performance panel for detailed interaction timeline analysis
• Web Vitals Chrome Extension — real-time INP measurement during manual testing sessions
• WebPageTest — INP testing with interaction simulation across multiple locations

The critical distinction: INP is primarily a Real User Monitoring (RUM) metric. Lab tests can simulate interactions, but INP scores in Google Search Console reflect actual user behavior patterns that cannot be fully replicated in synthetic tests.

Connecting INP to AI Search Visibility: Dageno AI

Optimizing INP improves user experience signals that contribute to page authority — and page authority is one of the predictors of AI citation eligibility. But once your technical performance is optimized, you still need a dedicated layer to know whether that optimization is translating into actual AI citations.

Dageno AI provides that measurement layer — tracking brand citation performance across ChatGPT, Perplexity, Google AI Overviews, Google AI Mode, Gemini, Claude, Grok, Microsoft Copilot, DeepSeek, and Qwen simultaneously from a single dashboard.

Where INP optimization addresses what happens when AI crawlers visit your site (JavaScript rendering quality, content accessibility), Dageno AI's knowledge graph integration addresses what AI platforms say about your brand when they do — ensuring that the content your optimized, well-performing pages contain is accurately characterized in AI-generated responses.

The connection is sequential: good INP → better AI crawler access → higher AI indexation probability → Dageno AI monitors whether indexed pages earn AI citations → knowledge graph alignment ensures those citations are accurate.

Pricing: Free plan available. Paid plans scale with prompt volume and monitoring frequency.

FAQ

Should I still care about FID after the switch to INP?
FID no longer affects Core Web Vital assessments in Google Search Console or ranking signals. Monitoring it is no longer necessary. However, since FID was always a subset of INP (the input delay component), optimizing INP will naturally address whatever FID issues may have existed.

Is INP a ranking factor?
Yes — INP is part of Core Web Vitals, which Google uses as a ranking signal within its page experience evaluation. Failing INP does not guarantee ranking drops, but Google has confirmed that Core Web Vitals are a tiebreaker signal when other ranking factors are similar between pages.

What causes poor INP scores?
Most poor INP scores trace to long JavaScript tasks on the main thread that delay event processing, heavy third-party scripts that interfere with interaction handling, or complex DOM updates that delay visual rendering. The same optimization that reduces these issues — code splitting, deferring non-critical JavaScript, reducing main thread blocking — also improves AI crawler content accessibility.

References

• Google Search Central – Introducing INP as Core Web Vital: Official Rationale for FID Replacement, "Better Evaluate Quality of User Experience" Justification
• web.dev – Interaction to Next Paint (INP): Complete Metric Definition, 200ms/500ms Thresholds, Input+Processing+Presentation Delay Components, 50-Interaction Exception Rule
• HTTP Archive / Chrome UX Report – Core Web Vitals Pass Rates: FID 90%+ Pass Rate vs INP 30%+ Mobile Failure Rate, Historical Trend Data
• Cloudflare – From Googlebot to GPTBot (May 2025): GPTBot 7.7% Crawler Market Share (+305%), PerplexityBot +157,490%, AI Bot Traffic Growth and JavaScript Rendering Limitations
• web.dev – Optimize INP: Long Task Reduction, Main Thread Unblocking, JavaScript Execution Optimization, Interaction-to-Frame Latency Diagnosis