Ad Fraud Statistics: Current Costs, Rates, and Industry Data

The numbers are worse than most people think – and they’re getting worse.

You’re probably here because you need to build a business case, benchmark your campaigns, or justify an investment in fraud prevention to your finance director. Maybe you’ve seen suspicious traffic patterns in your own data, or you’re trying to understand what “normal” fraud exposure looks like across the industry. Either way, you need reliable, authoritative figures – not vendor scare tactics or outdated blog posts.

This page consolidates the most recent ad fraud statistics, cost projections, and real-world case studies from authoritative sources. It’s designed to help UK media traders, designers, and agency teams understand the scale of the problem, quantify potential losses, and compare their own performance against industry benchmarks.

The Financial Scale of Ad Fraud

Global ad fraud cost reached $88 billion in 2023 and is projected to grow to $172 billion by 2028, according to Statista. That trajectory represents nearly a doubling of losses in just five years – a sobering reminder that fraud schemes are evolving faster than many prevention measures.

For context, 20–30% of digital ad spend is affected by fraud globally in 2025. That means roughly one in every five pounds you allocate to digital advertising could be going to bots, click farms, or spoofed inventory rather than real human audiences. If you’re managing a £500,000 annual digital budget, you’re potentially losing £100,000 to £150,000 to fraud. That’s not a rounding error – it’s the difference between hitting your targets and missing them entirely.

Historical data helps frame how persistent this problem has been. Back in 2015, AdAge reported that roughly one in every three dollars spent on digital advertising was thought to be lost to fraud – a rate exceeding 30%. While detection technology has improved, losses have grown in absolute terms simply because global ad spend has expanded so dramatically. A 2022 projection suggested that with programmatic advertising reaching $500 billion, roughly $100 billion of that would be lost to fraud. The percentage may have improved slightly, but the absolute cost has climbed relentlessly.

So ad fraud isn’t just a technical nuisance. It’s a structural tax on the entire digital advertising ecosystem – and you’re paying it whether you know it or not.

What Invalid Traffic Actually Looks Like

Invalid traffic is the industry term for impressions, clicks, and conversions that don’t come from genuine human users. Invalid traffic (IVT) accounts for 11% of global digital ad traffic, according to Statista. That’s the baseline “noise” in the system – traffic that shouldn’t be counted, billed, or optimised against.

But IVT rates vary significantly by channel and campaign type. Search campaigns face particularly acute fraud pressure. Search campaign click fraud rates range from 14% to 22%, with some studies suggesting up to 20% of paid search clicks are fraudulent. If you’re running high-volume PPC campaigns, that means one in five clicks might be generated by bots or competitors deliberately inflating your costs. When you’re paying £3–5 per click, that adds up quickly.

The bot problem extends across all digital channels. 24% of all ad clicks are bot-generated, and 40% of click fraud stems from bot networks specifically designed to mimic human behaviour. These aren’t simple scripts running on a single server – they’re sophisticated systems distributed across thousands of infected devices that can scroll, fill forms, and interact with content in ways designed to evade basic detection.

Adobe data cited in industry reports suggests that non-human traffic comprises 15% of digital traffic overall. That figure includes both malicious bots and legitimate automated systems (like search engine crawlers), but it underscores just how much of the web’s “activity” isn’t actually human. When you look at your analytics dashboard and see thousands of sessions, a significant portion of that traffic is automated systems, not potential customers.

Where Fraud Hits Hardest

Ad fraud isn’t distributed evenly across the globe. Statista identifies North America as the largest ad fraud market globally, followed by Asia-Pacific, Europe, Latin America, and the Middle East and Africa. The regional impact ranking matters because it tells you where the most sophisticated fraud operations are concentrated and where your campaigns face the highest risk.

Why does North America face the highest losses? Simple: it’s the largest digital advertising market, which makes it the most lucrative target for fraudsters. High CPMs and sophisticated programmatic infrastructure create more opportunities – and higher payouts – for fraud schemes. When a fraudster can generate $50 CPMs on premium North American inventory versus $5 CPMs elsewhere, the incentive to target that market becomes overwhelming.

Interestingly, in 2023, India surpassed the United States as the country with the highest ad fraud prevalence. The U.S. dropped to second place globally, highlighting how fraud activity is shifting toward emerging markets with rapid digital adoption and less mature fraud detection infrastructure. This shift reflects a broader pattern: as detection improves in established markets, fraudsters move to regions where defences are weaker and growth is faster.

Within the UK and Europe, fraud rates tend to be somewhat lower than in North America, but the problem is far from negligible. The global nature of programmatic advertising means that UK advertisers frequently buy inventory on international exchanges, exposing them to fraud originating anywhere in the world. Domain spoofing, geo-masking, and SDK manipulation allow fraudsters to disguise low-quality or non-existent traffic as premium UK inventory. You think you’re buying placements on quality UK publisher sites, but in reality, you’re getting spoofed inventory from data centres in Eastern Europe or Southeast Asia.

Industries Under Siege

Not all industries face the same level of fraud exposure. Certain verticals attract disproportionately high levels of click fraud, often because of high CPCs, intense local competition, and the ease of automating fraudulent activity. Industry-specific click fraud rates reveal significant variation: photography faces 65% fraud rates, pest control 62%, and locksmiths 53%.

These figures illustrate how local service businesses with high-value keywords become magnets for click fraud. Competitors may deliberately click on ads to exhaust budgets, and bot networks target these verticals because the traffic appears legitimate and the payouts are high. When a single click costs £8–12 and conversion rates are already low, even modest levels of click fraud can destroy profitability.

For agencies managing multi-vertical portfolios, this means fraud mitigation can’t be one-size-fits-all. High-risk verticals require more aggressive monitoring, stricter IP filtering, and closer scrutiny of traffic sources. A fraud prevention strategy that works for an e-commerce brand might be completely inadequate for a locksmith or pest control company operating in competitive local markets.

The Small Business Penalty

Larger advertisers with dedicated fraud prevention teams and sophisticated detection tools can limit their exposure. Small businesses, however, often lack those resources – and they pay the price. Small businesses lose up to 30% of ad budgets to fraud, a figure that can cripple marketing ROI and make digital advertising feel like a losing proposition.

When a local business with a £5,000 monthly ad budget loses £1,500 to bots and fraudulent clicks, the impact on growth can be severe. That’s £1,500 that could have funded genuine customer acquisition, creative development, or channel testing. Instead, it’s simply gone – stolen by systems designed to look just legitimate enough to slip past basic detection.

This dynamic creates a structural disadvantage. Larger brands can absorb fraud losses as a cost of doing business and invest in sophisticated mitigation tools. Smaller players often can’t, which means they either overpay dramatically for genuine traffic or exit digital channels altogether. The irony is bitter: the businesses that can least afford to waste money are the ones most vulnerable to fraud.

The Power of Accredited Partners

The good news is that working with accredited partners makes a measurable difference. Accredited partners reduce fraud rates from 8.99% to just 0.53% – a reduction of more than 90%. Let that sink in: choosing the right partners can cut your fraud exposure by an order of magnitude.

Accreditation typically involves certification from organisations like the Trustworthy Accountability Group (TAG), which sets standards for fraud prevention, brand safety, and transparency. Accredited vendors undergo regular audits and implement multi-layered fraud detection systems that catch invalid traffic before it’s billed. They maintain IP blacklists, monitor for suspicious patterns, and use behavioural analysis to identify bot activity that simpler systems miss.

For media traders and agencies, this underscores the importance of due diligence when selecting DSPs, SSPs, ad networks, and ad serving partners. A vendor’s accreditation status isn’t just a box to tick – it’s a direct indicator of fraud exposure and campaign quality. When you’re evaluating platforms, ask about TAG certification, request case studies showing fraud reduction, and demand transparent reporting on invalid traffic rates.

The cost difference between accredited and non-accredited partners may seem significant upfront, but the fraud savings typically dwarf any premium you pay. On a £1 million annual budget, reducing fraud from 9% to 0.5% saves roughly £85,000 per year. That’s often far more than the cost difference between premium and budget inventory sources.

How Fraud Actually Works

Understanding which fraud schemes are most common helps prioritise prevention efforts. The major categories include tactics that range from crude to sophisticated, and knowing how they operate makes detection much easier.

Click fraud remains the most visible type, where bots or human click farms repeatedly click on ads with no intent to engage. This inflates costs and distorts campaign metrics, making it harder to optimise for genuine performance. Modern click fraud operations use distributed networks and randomised behaviour patterns to avoid detection, making them significantly harder to spot than simple bot scripts.

Domain spoofing involves fraudsters disguising low-quality or non-existent websites as premium publishers, causing advertisers to pay premium CPMs for worthless inventory. This is particularly common in programmatic advertising, where inventory passes through multiple intermediaries before reaching the advertiser. A fraudster creates a fake website that claims to be The Guardian or BBC, lists it on an exchange, and collects premium rates for traffic that’s actually coming from a spam site or doesn’t exist at all.

Ad stacking places multiple ads in a single placement, with only the top ad visible to the user. All ads in the stack register impressions, even though only one could possibly be viewed. This scheme exploits weak viewability standards that only check whether an ad is within the viewport, not whether it’s actually visible to users. Advanced viewability measurement – such as the five-pixel standard that requires all four corners and the centre of an ad to be visible – can reduce ad stacking fraud by 90–95%.

Pixel stuffing renders ads in tiny, invisible 1×1 pixel placements that users never see, yet count as delivered impressions. Like ad stacking, this scheme exploits weak viewability standards and can generate massive volumes of fraudulent impressions at minimal cost.

Geo-masking disguises the geographical location of traffic, making it appear to originate from high-value target regions when it actually comes from low-cost, low-quality sources. This is especially problematic for campaigns with strict geographic targeting. You think you’re reaching London-based professionals, but the traffic is actually coming from click farms in Bangladesh or Vietnam.

Bot networks and bot farms have become increasingly sophisticated. Modern bots can scroll, hover, click, fill forms, and even mimic mouse movement patterns to evade detection. Some botnets operate across millions of infected devices, while bot farms run in data centres with dedicated infrastructure designed to simulate human behaviour at scale. The most advanced systems can pass basic CAPTCHA tests, maintain consistent session times, and interact with content in ways that look entirely human to simple detection algorithms.

For a more detailed explanation of these fraud types and how they work, see our guide on ad fraud detection and prevention.

Case Studies: Fraud in Action

Fraud statistics are abstract until you see how schemes actually operate. A few notable cases illustrate the scale and sophistication of ad fraud operations and show just how much money is at stake.

“We Purchase Apps” involved fraudsters buying legitimate apps and then injecting bot traffic to generate fraudulent ad revenue. By acquiring apps with existing user bases, they could blend fake traffic with real usage, making detection more difficult. The operation ran for months before being discovered, generating millions in fraudulent revenue.

HummingBad was a malware operation by the Yingmob group that infected millions of Android devices. The malware generated fraudulent ad impressions and clicks, netting the operators millions in revenue before being shut down. The sophistication was remarkable: HummingBad could root devices, install persistent malware, and simulate genuine user behaviour well enough to evade most detection systems.

Chamois was discovered by Google and represented one of the largest mobile malware campaigns at the time, pre-installing malicious code on devices before they reached users. The operation infected devices at the manufacturing stage, meaning users received compromised phones straight out of the box.

Thai WeChat Click Farm was a physical operation employing hundreds of people to manually click ads and engage with content. While less scalable than bots, human click farms can bypass detection systems designed to catch automated behaviour. Photos from the operation showed rooms filled with workers, each managing dozens of phones simultaneously, manually clicking ads and engaging with content to generate fraudulent revenue.

Zirconium combined malvertising with false redirects, sending users to fraudulent pages while generating billable impressions for the original ad placement. The operation used sophisticated social engineering and technical exploits to compromise legitimate advertising platforms, making it exceptionally difficult to detect and shut down.

These cases aren’t historical curiosities. Similar schemes operate today, often at even greater scale and with more sophisticated evasion techniques. The fraudsters learn from each takedown, adapting their methods to avoid the detection systems that caught their predecessors.

Mobile and Regional Fraud Dynamics

Mobile advertising faces unique fraud challenges that stem from the fragmented nature of app ecosystems and the difficulty of verifying mobile traffic. Mobile ad fraud has become increasingly sophisticated, with SDK spoofing, fake app installs, and in-app fraud growing in prevalence. High mobile adoption rates – especially in Asia-Pacific – create lucrative opportunities for fraudsters who can exploit less mature detection infrastructure.

SDK spoofing manipulates the software development kit in mobile apps to generate fake ad engagements, such as clicks or installs. Because the fraud occurs at the SDK level, it’s harder to detect through traditional traffic analysis. The fraudster essentially hijacks the communication between the app and the ad network, making fake events appear to come from legitimate app activity.

The Asia-Pacific region’s rapid mobile growth has made it a hotspot for fraud. High mobile penetration, diverse device ecosystems, and fragmented regulation create an environment where fraudsters can operate with relative impunity. For UK advertisers running mobile campaigns that include APAC inventory – or for brands with international reach – understanding these regional dynamics is critical. Fraud prevention measures need to account for mobile-specific threats and regional variations in fraud tactics.

The programmatic advertising ecosystem’s automated nature has increased vulnerabilities to ad fraud despite offering efficiency and scale. The speed and complexity of programmatic transactions create opportunities for fraudsters to inject fake inventory, spoof premium placements, and exploit gaps in verification systems. When buying decisions are made in milliseconds across dozens of intermediaries, it becomes exponentially harder to verify that each impression is genuine.

Emerging Threats on the Horizon

Ad fraud isn’t static. As detection improves, fraudsters adapt. Several emerging trends are worth watching because they represent the next generation of fraud schemes – and they’re significantly harder to detect than current methods.

AI-powered fraud is the most significant new threat. Generative AI can create realistic fake user profiles, generate synthetic traffic patterns, and even produce deepfake content for influencer fraud. These tools lower the barrier to entry for fraud and make detection exponentially harder. Where it once took significant technical skill to create convincing fake traffic, AI tools can now generate sophisticated fraud operations with minimal expertise.

Connected TV (CTV) and OTT fraud is growing as advertising budgets shift toward streaming platforms. CTV fraud includes fake app installs, spoofed device IDs, and server-side ad insertion fraud. Because CTV measurement is less mature than web or mobile, detection lags behind. Fraudsters are exploiting this gap, creating sophisticated schemes that take advantage of the unique technical architecture of streaming platforms.

Deepfakes for influencer fraud involve using AI-generated video or images to impersonate influencers or create entirely fictitious personas. Brands can pay for endorsements that never actually happened, delivered to audiences that don’t exist. This scheme is particularly insidious because it combines fraud with reputational risk – brands might unknowingly associate themselves with fake influencers or fraudulent content.

Fraud-as-a-Service platforms have commoditised fraud, making sophisticated tools available to anyone willing to pay. These platforms offer click fraud, traffic generation, and fake engagement as productised services, with customer support and tutorials included. The professionalisation of fraud means that even unsophisticated actors can now launch effective fraud operations, dramatically expanding the threat landscape.

For more on how advertising technology is evolving to address these threats, see our article on top trends in new advertising technology.

The Ripple Effects on Publishers and Advertisers

Ad fraud doesn’t just waste advertiser budgets – it also harms publishers who depend on advertising revenue to fund content creation. Fraudulent traffic distorts performance metrics, making it harder for publishers to demonstrate value to advertisers. When fraud is detected after the fact, publishers risk revenue clawbacks, damaged reputations, and loss of premium advertising partnerships. Quality publishers get lumped in with fraudulent inventory sources, driving down rates across the board.

For advertisers, the consequences extend beyond wasted spend. Inflated costs, skewed analytics, and poor ROI make it difficult to forecast accurately, optimise campaigns, or justify continued investment in digital channels. When your data is polluted by bot traffic, every decision you make – from creative testing to audience targeting – is based on false signals. You might abandon a high-performing creative because bots didn’t engage with it, or double down on a channel that’s delivering fraudulent conversions.

This creates a vicious cycle. Advertisers lose trust in digital channels, which reduces demand for inventory, which pressures publishers to accept lower-quality demand sources, which increases fraud exposure even further. The entire ecosystem suffers when fraud goes unchecked, making it harder for legitimate players to operate profitably.

What Actually Works: Multi-Level Defence

Multi-level blocking has been identified as the most effective mitigation strategy against the full range of ad fraud risks. This approach combines multiple detection layers, each designed to catch different types of fraud at different points in the campaign lifecycle.

Traffic-level filtering blocks known bot IPs, data centre traffic, and suspicious user agents before ads are served. This catches the simplest fraud but misses sophisticated operations that use residential IPs and realistic user agents. Behavioural analysis uses machine learning to identify patterns indicative of fraud, such as unrealistic click speeds, impossible navigation paths, or engagement patterns that don’t match human behaviour. Viewability measurement ensures ads meet strict standards for visibility, reducing pixel stuffing and ad stacking.

Post-impression analysis detects anomalies in conversion rates, time-to-conversion, and engagement patterns that suggest fraud even when the initial traffic looks legitimate. Third-party verification from accredited partners provides independent validation of traffic quality, adding an external check on your internal systems.

For agencies and in-house teams, this means fraud prevention can’t be outsourced to a single tool or vendor. It requires a coordinated approach across media planning, campaign setup, real-time monitoring, and post-campaign analysis. You need to think about fraud at every stage of the campaign lifecycle, from inventory selection through to conversion attribution.

Our guide on ad fraud prevention offers practical steps for implementing multi-layered defences, from setting IP blacklists to monitoring copyright theft and configuring custom alerts that notify you immediately when metrics deviate from expected patterns.

Supply Chain Transparency and Ads.txt

One of the simplest yet most effective fraud prevention measures is ads.txt, a standard developed by the IAB to combat domain spoofing. Ads.txt files let publishers declare which vendors are authorised to sell their inventory, making it much harder for fraudsters to impersonate premium publishers. When you check a publisher’s ads.txt file and verify that the SSP you’re buying through is listed, you dramatically reduce the risk of domain spoofing.

Despite its simplicity, ads.txt adoption remains incomplete. Some publishers haven’t implemented it properly, and some advertisers don’t enforce it strictly enough. But when properly deployed, ads.txt provides a clear record of the authorised supply chain, reducing unauthorised reselling and improving transparency across the ecosystem.

For media traders, checking ads.txt files before buying inventory should be standard practice – as routine as checking viewability benchmarks or audience demographics. For publishers, maintaining an up-to-date ads.txt file protects your inventory value and builds trust with advertisers who take fraud prevention seriously.

Related transparency efforts include supply-side platforms that offer detailed reporting on inventory sources, fill rates, and traffic quality. The best SSPs provide granular data on where inventory originates, how it’s categorised, and what verification systems have been applied. This transparency makes it possible to spot anomalies and verify that you’re actually getting what you’re paying for.

Balancing Fraud Prevention with Privacy

The shift toward privacy-first advertising complicates fraud detection in ways that aren’t immediately obvious. Third-party tracking has historically been a key tool for identifying suspicious patterns across sites, but GDPR, browser restrictions, and the evolving treatment of third-party cookies limit what data can be collected and how it can be used.

This creates a genuine tension: fraud prevention often relies on the same tracking mechanisms that privacy regulations restrict. The solution isn’t to abandon privacy protections – it’s to develop new detection methods that work within privacy constraints and don’t compromise user trust.

First-party data offers one path forward. By collecting data directly from your own properties and audiences, you gain greater data accuracy and reliability while maintaining privacy compliance. First-party data also reduces dependency on third-party sources, which are more vulnerable to fraud because they aggregate data from multiple sources with varying quality standards.

Contextual targeting – aligning ads with the theme of the content rather than user behaviour – is another privacy-friendly approach that’s experiencing renewed interest. While contextual targeting doesn’t eliminate fraud, it reduces exposure to audience-based fraud schemes that rely on spoofed user profiles and fake behavioural data. Contextual targeting is growing due to tightening privacy regulations, but it also offers fraud reduction benefits that make it attractive even in the absence of regulatory pressure.

Practical Steps for Your Campaigns

If you’re managing digital advertising in the UK, here’s what these statistics should tell you about your immediate priorities:

Budget conservatively for fraud losses. If you’re not actively monitoring for fraud with sophisticated detection tools, assume at least 10–15% of your spend is going to invalid traffic. For high-risk verticals or untested inventory sources, that figure could be 20–30% or higher. Build this into your forecasts and set aside contingency budget for fraud mitigation tools.

Prioritise accredited partners. The difference between an 8.99% fraud rate and a 0.53% fraud rate is enormous when you’re managing six- or seven-figure budgets. Vet your DSPs, SSPs, ad networks, and creative platforms for fraud prevention capabilities and industry certifications. Ask specific questions about TAG certification, fraud detection methodologies, and historical fraud rates.

Implement multi-layered detection. Don’t rely on a single tool or metric. Combine traffic filtering, behavioural analysis, viewability measurement, and post-impression analysis to catch fraud at multiple points in the campaign lifecycle. Each layer catches different types of fraud, and the combination is far more effective than any single approach.

Monitor metrics for anomalies. Sudden spikes in CTR, traffic from unexpected geographies, or conversion rates that don’t match your historical benchmarks are all red flags. Set up custom alerts so you’re notified immediately when metrics deviate from expected ranges. The faster you catch fraud, the less money you lose.

Audit your campaigns regularly. Fraud detection isn’t a one-time setup. Fraudsters adapt, so your defences need to evolve too. Quarterly audits of traffic sources, performance metrics, and vendor practices help catch emerging threats before they become expensive problems. Schedule these audits in advance and treat them as non-negotiable maintenance.

For brands building rich media and programmatic campaigns, lightweight creatives that load quickly also help with fraud detection. Faster load times make it easier to spot suspicious patterns in traffic behaviour, and creatives that require genuine engagement – rather than just passive impressions – are harder for bots to fake convincingly. Learn more about how engagement metrics can combat poor visibility.

Making the Business Case

If you’re trying to justify investment in fraud prevention to your finance director or senior leadership, these statistics provide the foundation for a compelling business case that focuses on hard ROI rather than abstract risk reduction.

Start by quantifying current losses. Take your annual digital ad spend and multiply by 10–20% (or higher for high-risk verticals). That’s a conservative estimate of what you’re losing to fraud right now. Present this as a line item: “Current estimated fraud losses: £100,000 per year.” That makes the problem concrete and immediate.

Next, calculate the ROI of prevention. If accredited partners reduce fraud rates from 9% to 0.5%, the savings on a £1 million budget are roughly £85,000 per year. That’s often far more than the cost of verification services or the premium you pay for accredited inventory. Show the payback period – typically measured in months, not years – and emphasise that this is a recurring annual saving.

Frame fraud as a data quality issue, not just a cost problem. Fraud doesn’t just waste money – it pollutes your analytics, making every other decision (creative testing, audience targeting, channel allocation) less effective. Cleaning up your data improves ROI across your entire marketing stack, creating compounding benefits that extend well beyond the direct fraud savings.

Highlight competitive risk. If your competitors are actively managing fraud and you’re not, they’re getting better data, better performance, and better ROI. That’s a competitive disadvantage that compounds over time. They can afford to bid more aggressively, test more creatives, and expand into new channels because they’re not subsidising fraudsters with 20% of their budget.

Reference case studies that show real-world outcomes. Real examples of brands that reduced fraud and improved performance make the business case tangible rather than theoretical. The data in this article provides plenty of material for building that narrative.

Where to Start

Ad fraud is pervasive, costly, and constantly evolving – but it’s not insurmountable. With the right data, tools, and processes, you can significantly reduce exposure and improve campaign performance without needing a dedicated fraud team or unlimited budget.

Start by benchmarking your current fraud levels. Review your traffic sources, compare your performance metrics to the industry standards documented here, and identify where you’re most vulnerable. Look for the red flags: unusually high CTRs in specific channels, traffic spikes from unexpected regions, conversion patterns that don’t match your historical data.

Then prioritise the highest-impact prevention measures: switching to accredited partners, implementing multi-layered detection, and establishing regular audit processes. You don’t need to fix everything at once, but you do need to stop the biggest leaks first. Focus on the 20% of actions that will eliminate 80% of your fraud exposure.

If you’re building programmatic campaigns and want built-in fraud protection, NEXD’s Campaign Manager includes ad serving with no extra fees and uses lightweight, GPU-rendered creatives that make fraudulent traffic easier to detect. The platform’s five-pixel viewability measurement ensures ads meet strict visibility standards, reducing ad stacking and pixel stuffing by up to 95%. Faster load times help identify patterns indicative of bot traffic, and the Campaign Manager provides per-asset interaction data so you can see exactly which elements are driving genuine engagement.

Sign up to NEXD Campaign Manager for a free 14-day trial and see how faster-loading, engagement-driven creatives can help you reach real audiences while keeping fraud at bay.

For more foundational knowledge on how fraud schemes work and how to detect them, read our complete guide to ad fraud explained. And if you’re ready to implement prevention strategies immediately, our guide to ad fraud prevention offers actionable steps you can take today.