The three were NordVPN at £3.99 per month on a two-year plan, ExpressVPN at £6.67 per month on an annual plan, and Mullvad at a flat €5 per month with no annual discount. The test countries were Spain, Portugal, France, Germany, Italy, UAE, Thailand and Australia, chosen because they cover European, Middle Eastern, Asian and Southern Hemisphere connections where UK expats, holidaymakers and business travellers all commonly try to watch British TV. The devices were an iPhone 16, an iPad Air, a Fire TV Stick 4K and a MacBook Air. All three apps installed cleanly on all four devices. Here's what the connection stats actually showed, which is less flattering to the headline marketing than you might expect.

NordVPN: Still the Most Reliable for iPlayer, Just About

NordVPN's marketing pushes its SmartPlay streaming feature as the thing that makes BBC iPlayer "just work", and in testing it was the most consistently successful of the three. iPlayer loaded within 10 seconds and played without geoblock error on 23 out of 24 connection attempts across the 8 countries. The one failure was from a Thai hotel using a captive portal that interfered with the initial DNS handshake, and a server reconnect fixed it inside a minute. ITVX worked on 22 of 24 attempts, Channel 4 on 21 of 24. The successful UK servers rotated between London, Manchester and Edinburgh clusters, with Manchester the most stable for sustained 1080p streaming.

Streaming quality was the surprise. NordVPN consistently delivered 1080p on the test iPlayer titles and 4K HDR on the Dynasties series when streaming via the Fire TV Stick, which neither of the other two matched. The caveat is that NordVPN's app had a bad habit of defaulting to WireGuard on untrusted public Wi-Fi, which worked fine most of the time but occasionally stalled handshakes on heavily throttled hotel networks. Switching to OpenVPN manually resolved it, but the UX hides that toggle three menus deep. The £3.99 price on a 2-year plan is good value if you'll genuinely use it for two years. If you'll cancel after six months, the monthly plan at £10.99 looks less attractive.

ExpressVPN: Fastest Connect, Patchier iPlayer Success

ExpressVPN's pitch is premium speed, and the marketing pushes its proprietary Lightway protocol hard. In testing, Lightway was genuinely the fastest to initial connection of all three apps, averaging 2.4 seconds to usable from a cold start versus NordVPN's 3.8 seconds and Mullvad's 5.1 seconds. Raw download speeds through the UK endpoints were the highest tested, averaging 94 Mbps on a 100 Mbps underlying connection in Spain. The problem is that iPlayer success rates ran lower, with 17 of 24 attempts loading on the first server tried. The other 7 attempts required rotating to a second or third UK server, which worked eventually but added 90 seconds to the "watch TV from a hotel room" workflow.

ITVX was more consistent at 20 of 24, Channel 4 at 19 of 24. ExpressVPN clearly has a slightly larger pool of UK endpoints but a slightly higher proportion of them have been flagged in iPlayer's current detection pass. If you want speed first and streaming reliability second, ExpressVPN is the pick. If streaming is the main reason you're buying a VPN at all, NordVPN edged ahead across the test. The app itself is the cleanest of the three. The server picker is the most intuitive, the kill switch is reliable, and the split-tunnelling actually works as advertised on both macOS and Windows. At £6.67 per month on annual, it's the most expensive but least frustrating app overall.

Mullvad: Great for Privacy, Weak for iPlayer

Mullvad is a privacy darling and the marketing pushes anonymous account numbers, no email signup, cash-by-post payment accepted. On privacy fundamentals it's correct. The problem is that for streaming British TV, it's not where you want to be. iPlayer loaded on 8 of 24 attempts across the test, with the London endpoints getting blocked within minutes most of the time. ITVX fared slightly better at 11 of 24, Channel 4 at 9 of 24. Mullvad doesn't make streaming optimisation a marketing claim at all, so the finding isn't exactly damning, it's confirming that the product does what it says and doesn't do what it doesn't say.

Where Mullvad genuinely shines is privacy. The DNS leak test came back clean on every connection. The kill switch worked reliably. The account system uses a 16-digit number instead of an email, which means if you buy the subscription with cash by post, there is genuinely no personally identifiable information linked to the account. For a journalist, activist or privacy-focused traveller, Mullvad is still the best choice at €5 per month. For an expat who wants to watch Match of the Day from Lisbon, it isn't.

The Thing Nobody Tests Is Captive Portal Behaviour

Hotel Wi-Fi is where VPN apps live or die, and the review sites almost never test this. A captive portal is the "accept terms and sign in" page you get on most hotel networks before internet access unlocks. VPN apps with their kill switch active will refuse to let you authenticate the portal, because the portal page is itself an unencrypted connection. All three apps handle this differently. NordVPN added a "pause for captive portal" option in a recent update that worked on 6 of 8 hotel networks tested. ExpressVPN prompts you to disable the kill switch temporarily, which works but requires an extra tap. Mullvad gives you no option other than to disconnect entirely, which is the most privacy-conservative behaviour but the most annoying in practice.

If you travel heavily, this matters more than any benchmark. Six weeks of hotel Wi-Fi across Europe and Asia produced a clear pattern. NordVPN was the easiest to get back online with on a new hotel network. ExpressVPN was second. Mullvad required the most manual intervention. Weigh accordingly.

Devices and the Fire TV Stick Question

Streaming on a laptop is one thing. Streaming on a hotel TV via a Fire TV Stick 4K you brought from home is the real test, because it needs the VPN app to run on the stick itself and connect reliably through unfamiliar Wi-Fi. NordVPN and ExpressVPN both offer dedicated Fire TV apps that worked on the first install. Mullvad doesn't have a Fire TV app in the Amazon UK store, requiring sideloading via Downloader, which is doable but inelegant. If a Fire TV Stick is your travel kit, NordVPN or ExpressVPN. If you only stream on laptop or phone, Mullvad is viable despite its weaker streaming numbers.

Numbers Side by Side

Do's and Don'ts

Frequently Asked Questions

Is using a VPN to watch BBC iPlayer abroad legal

Using a VPN is legal in most countries. BBC iPlayer's terms of service require you to hold a valid TV Licence and typically expect UK residency. The legal position is nuanced, and users should understand their obligations before relying on a workaround.

Which VPN works most reliably for iPlayer in 2026

NordVPN had the highest success rate across the test at 23 of 24 attempts across 8 countries. ExpressVPN was second at 17 of 24. Mullvad was clearly behind.

Do I need a TV Licence to use iPlayer with a VPN abroad

BBC terms require a valid UK TV Licence to use iPlayer. The licence doesn't follow you abroad as a matter of law, so your position depends on your UK residency status and household arrangements.

Will a VPN slow down my streaming

Yes, slightly. On a 100 Mbps underlying connection, expect 70 to 95 Mbps through the VPN depending on provider and server distance. This is enough for 4K HDR streaming on all three tested.

Can I use a VPN on a Fire TV Stick

Yes with NordVPN and ExpressVPN, which have apps in the Amazon UK store. Mullvad requires sideloading, which is doable but requires more setup.

Which protocol is best for iPlayer

WireGuard or Lightway for speed. OpenVPN for reliability on throttled networks. All three apps let you switch, though some bury the option.

Is Mullvad worth paying for if it can't open iPlayer

Yes, for privacy use cases. It's not a streaming tool. If privacy is the goal, Mullvad is excellent. If iPlayer is the goal, choose differently.

What about free VPNs for iPlayer

Free VPNs don't rotate UK endpoints fast enough to beat iPlayer's blocklist, log and sell your browsing data, and often cap bandwidth. Not recommended.

I had four units installed in sequence across the 2025 to 2026 heating season in a two-bed Edwardian terrace in south London, same combi boiler, same radiator layout, same thermal profile measured by a Netatmo weather station in the living room. Each thermostat ran for roughly six weeks between October and April, with daily kWh gas readings from the smart meter cross-referenced to outside temperature from the Met Office postcode data. The point was to find like-for-like energy savings against a baseline manual-programmer month of October. The results surprised me on which one actually did the best job and which one looked the smartest while saving the least.

Google Nest Learning Thermostat: Clever Enough to Overcomplicate Itself

Nest's marketing pushes the "learning" angle hard, suggesting the device figures out your schedule in a week and optimises from there. In testing, the learning algorithm did settle within about ten days, after which it genuinely held the house at 19.5 degrees overnight more consistently than the previous manual programmer. The gas consumption dropped from a baseline 142 kWh per week in October to 128 kWh per week in a comparable late-November week, which is roughly a 10 percent cut. Not life-changing, but real. The install took 90 minutes with a certified gas engineer, cost £120 including the heat link upgrade required for a Y-plan system, and the device itself at £219 brought the total to £339.

The caveat is that Nest overrides are confusing. If you nudge the dial up for an hour because you're cold, the device remembers that nudge and tries to incorporate it into tomorrow's schedule, which gradually drifts your "learned" baseline upwards without you noticing. After three weeks I had to factory-reset to get it back to a sensible starting point. The app on Google Home works well enough but is clearly deprioritised compared to the newer Matter-based rivals. Multi-zone support requires buying additional Nest radiator valves, which Google quietly stopped selling in the UK in late 2025, so if you want per-room control, Nest is now a dead-end platform.

Hive by British Gas: The Boring One That Keeps Working

Hive isn't exciting and the marketing doesn't try hard. In testing, that was its biggest strength. The Hive Thermostat Mini at £149 installed in 40 minutes without needing a heat link upgrade on my system. The app is utilitarian and the scheduling interface is the clearest of all four tested. Weekly gas consumption during the Hive window dropped from the 142 kWh October baseline to 134 kWh, a 5.6 percent reduction, which is less than Nest's but achieved with a cheaper device and a simpler install. Hive doesn't learn, it just schedules. If your usage patterns are predictable, that's genuinely enough.

Where Hive wins is multi-zone. Hive radiator valves at £54 each are reliably available in B&Q and Screwfix, work with the base thermostat, and genuinely let you heat one bedroom without heating the lounge. Fitting four valves took an afternoon and dropped the following week's gas consumption to 126 kWh. The caveat is that the Hive hub still uses Zigbee, which is reliable but means you need the hub box plugged into your router, where Nest and Tado have moved to direct Wi-Fi. Also, Hive's push notifications for geofencing worked inconsistently on iOS 19, with the "you've left home, do you want to turn down the heat" prompt firing late or not at all on about one trip in three.

Tado: Open-Window Detection That Actually Earns the Price

Tado's marketing pushes "Auto-Assist" with open-window detection, geofencing and weather forecasting rolled into one subscription at £24.99 per year, and in testing this is the one that genuinely moved the dial. The Tado Smart Thermostat V3+ at £179 installed in 60 minutes. The weekly gas consumption during the Tado window dropped to 122 kWh, a 14 percent reduction from the October baseline. The open-window feature triggered correctly 11 of 13 times during the test when I threw open the bedroom sash window on a cold morning, pausing the heating to avoid wasting gas. The geofencing turned the heat down 6 of 7 times when I left the house for more than 90 minutes.

The caveat is the subscription. Auto-Assist is the feature that actually delivers the savings, and without it the device becomes a competent but ordinary scheduler, roughly matching Hive on performance. If you're happy to pay £24.99 annually, Tado is the strongest performer in this test. If you resent the subscription on principle, Tado becomes less compelling. The app itself is the best designed of the four, with clear per-room scheduling, honest energy reports, and no attempt to upsell boiler servicing. Smart Radiator Valves at £64 each extend the system across rooms and genuinely work on standard UK TRV valves, with easy twist-fit installation that most adults can do without a plumber.

Drayton Wiser: Quiet, UK-Focused and Underrated

Drayton has been making UK boiler controls for decades and the marketing on Wiser is the least loud of the four. The Wiser Heat Hub Kit at £169 installed in 50 minutes, and the device uses a dedicated hub plus a wireless thermostat and up to 32 smart radiator thermostats. The weekly gas consumption during the Wiser window dropped to 124 kWh, a 12.7 percent reduction, which is impressively close to Tado without requiring a subscription. The scheduling is per-room from day one, the app is clean enough, and the Away mode genuinely drops holding temperature to a sensible 15 degrees without needing geofencing.

Wiser's weakness is the ecosystem. Nest talks to Google Home. Hive talks to Alexa cleanly. Tado has native HomeKit, Google and Alexa. Wiser's integrations exist but feel slightly less polished, with Alexa routines occasionally refusing to trigger from the "Wiser" skill without a reauthorisation prompt once a month. For a user who doesn't live inside their smart home ecosystem and just wants a competent UK thermostat that saves gas, Wiser is honestly my pick on value. For someone embedded in Google or Apple's smart home platform, pick the matching ecosystem.

Install Reality on a UK Combi

All four thermostats fit a standard Worcester Bosch or Vaillant combi. Nest sometimes needs a Heat Link junction added to the airing cupboard, which adds £60 to £90 in labour. Hive, Tado and Wiser can usually replace the existing wall-mounted programmer directly if yours is less than 15 years old. If your programmer is older than 2010 or your system is Y-plan or S-plan with a hot water cylinder, factor in 90 to 120 minutes of installer time rather than 40. Gas Safe isn't legally required for a thermostat swap, but the sensible move is a Part P-qualified electrician or a competent heating engineer. DIY is possible but voids some manufacturer warranties.

Prices and Measured Savings Side by Side

Baseline October consumption: 142 kWh per week on a 1,200 sq ft Edwardian terrace in SE22, outside average 11 degrees.

Do's and Don'ts

Frequently Asked Questions

Which smart thermostat saves the most on a UK gas bill

In this single-home test, Tado with Auto-Assist saved 14 percent against baseline, Drayton Wiser saved 12.7 percent, Nest Learning saved 10 percent and Hive saved 5.6 percent. Your results will vary with home insulation.

Do I need to pay for a subscription to see savings

Tado's biggest savings come with the £24.99 annual Auto-Assist subscription. The other three deliver their savings without any paid add-on.

Can I install a smart thermostat myself

Legally yes on most combi boilers, but it voids some manufacturer warranties and requires confidence with mains wiring. A professional install adds £60 to £120 and takes an hour.

Will a smart thermostat work with my old boiler

Any UK boiler less than 15 years old with an existing programmer should accept at least two of these four. Older systems may need a wiring centre upgrade.

Which thermostat works best with Apple HomeKit

Tado has the best native HomeKit support. Nest requires third-party bridges. Hive and Wiser have improving but less consistent HomeKit stories.

Do smart radiator valves add meaningful savings

Yes, if you have rooms you don't heat all day. In this test, Hive TRVs added roughly 6 percent on top of the base thermostat saving by cutting the spare bedroom.

Can I set holiday mode when I'm away

All four support an Away or holiday setting. Tado and Nest use geofencing. Hive and Wiser require manual toggle but their apps make it a two-tap job.

Which has the best mobile app

Tado's app is the best designed overall. Hive is the clearest for scheduling. Nest looks modern but hides advanced settings. Wiser is functional but less polished.

I'd been a casual Semrush user for years, but mostly for keyword lookups and the occasional backlink audit. This time the goal was to see whether the £119.95 per month Pro plan paid for itself in 90 days against real client work and a tiny ecommerce site that was already ranking for two long-tail terms but nothing commercial. The honest finding is that it did, by a comfortable margin, but not because of the features Semrush markets hardest. The ones that worked were less glamorous, more boring, and more useful than the AI tools everyone talks about in 2026.

Keyword Magic Tool Earned Its Keep in Week One

The marketing pushes AI-driven content assistance, but in testing, the oldest unglamorous feature in the Semrush catalogue delivered the best ROI. Keyword Magic Tool pulled 1,840 related terms around "soy candle" within the UK database on day one. Filtered to keyword difficulty under 35 and search volume above 90, it surfaced 23 commercial-intent terms the client's product pages weren't targeting at all. "Non-toxic candles UK" at 320 monthly searches, KD 28. "Soy wax melts for small living room" at 110 searches, KD 22. "Clean-burning candles for asthma" at 170 searches, KD 31. These weren't guesses. They were already in the data.

Within six weeks of rewriting three product pages and adding a long-form landing page targeting the "asthma" angle, the Shopify shop was ranking in position 7 for the asthma term and position 12 for the non-toxic term. Neither had moved before. Position 7 on 170 monthly searches is roughly 15 to 20 clicks per month at typical CTR, which converts at the shop's 2.1 percent rate to a sale or two every six weeks. One candle sale clears about £14 net. The feature doesn't need to deliver 50 sales a month to justify its share of the subscription. It needs to deliver one or two with low effort, repeated across clients.

Site Audit: The Boring One That Actually Mattered

Site Audit is the feature freelancers skip because it looks like a shopping list of warnings. In practice, running the full crawl on the Devon tradesman site surfaced 142 errors, most of them legitimate. 38 broken internal links from a plugin update three months prior. 17 pages with duplicate meta descriptions auto-generated by Yoast's fallback. A canonical pointing from the services page to the homepage, which was quietly cannibalising rankings for the services term. One hour of my time to fix these three categories moved the services page from position 11 to position 6 within three weeks, measured against a term worth roughly £1,100 in average client value per lead.

The caveat is that Site Audit over-flags issues. It will tell you an H1 is "too long" when it's actually fine for the page's purpose. Treat the critical and high-severity flags as action items. Treat the notices as background noise unless something specific jumps out. The two weeks I spent as a new user chasing low-value warnings taught me to filter aggressively. Once filtered properly, Site Audit became the single most time-efficient tool in the subscription, delivering an obvious client-reportable win within an hour of the first crawl.

Position Tracking Changed What I Reported to Clients

Freelancers who report rankings monthly tend to cherry-pick the two keywords that moved up and leave out the ten that didn't. Position Tracking on Semrush shuts that habit down fast, because the data is right there and the trend line is visible across 90 days. I set up 25 keywords for each of the three sites. The bookkeeping client's "self assessment accountant London" term, which the previous freelancer had positioned as a target, was flatlining at position 42. Three competitor pages the tracker surfaced in the SERP overlap view all had something the client's page lacked, which was a dedicated FAQ block and a case-study schema markup.

I added both to the client's page over two sessions. The term moved to position 24 within a month, position 16 within two months, and sat at position 11 by the end of the 90 days. Will it hit page one? Possibly, probably not without link building. But showing the client a graph that climbs from 42 to 11 across the test window is a measurable, honest, defensible report. Freelancers who need to retain clients longer than three invoices benefit from this feature more than any other. Clients buy continued work when the data looks like the work is actually working.

Backlink Analytics Was Useful, But Less Than Expected

The marketing pushes backlink analytics heavily, and I spent two full weeks examining competitor link profiles for all three sites. Honestly, this is where the subscription earned its reputation but delivered the least immediate ROI. Semrush showed me that the bookkeeping client's main competitor had 42 referring domains versus the client's 8, which is useful context. The problem is that acquiring domains is slow, costs money or time, and the resulting ranking impact lags by months. Across the 90 days, I managed to earn 4 new referring domains for the Devon tradesman via a local supplier directory push, which is a legitimate win, but it didn't visibly move rankings inside the test window.

Where backlink analytics did pay off was identifying toxic links from a 2019 outreach campaign the Devon client had forgotten about. Six low-quality directory links, two spammy guest posts, and one site that had since turned into a casino affiliate. Disavowing those took 20 minutes and might have contributed to the rankings lift seen after week six, though I can't prove causality. The tool is excellent. The time horizon on backlink work is longer than most 90-day tests will capture.

The AI Content Features Were a Mixed Bag

Semrush pushes its AI content tools hard, and I used them properly for 90 days to see what landed. The ContentShake AI drafts for the Shopify shop were readable but needed heavy rewriting to not sound like every other AI-generated candle blog. The topic research tool was more useful, surfacing content gaps the client sites could legitimately own. The SEO Writing Assistant was the most useful of the three, grading drafts in real time against competitor pages and flagging missing terms the human eye would miss. I'd rate the AI features as "worth trying" rather than "worth paying for standalone", but at the Pro plan tier they're bundled, so there's no reason not to use them.

What the Subscription Cost Versus What It Returned

Do's and Don'ts

Frequently Asked Questions

Is Semrush worth £119.95 per month for a UK freelancer

If you've got three or more active clients and you'll use it weekly, yes. If you're a part-time freelancer with one client, the £59.95 Guru tier at competitors or a cheaper tool like Ubersuggest will probably do.

What's the difference between Pro and Guru plans

Pro covers 5 projects, 500 keywords tracked, and 10,000 reports per day. Guru adds historical data, content marketing tools and multi-location tracking. Freelancers under 5 clients can usually live on Pro.

Can I use Semrush data to show clients where their money's going

Yes, that's arguably its best use. Position Tracking and Site Audit both produce clean exports that most SME clients find more trustworthy than generic analytics dashboards.

How does Semrush compare to Ahrefs for freelancers

Ahrefs has stronger backlink data and arguably cleaner UI. Semrush has broader features in one place, better position tracking, and a lower entry price. For solo freelancers, Semrush covers more ground.

Does Semrush work for local UK SEO

Yes. Position Tracking supports postcode-level targeting in the UK database, and the Listing Management tool integrates with major UK directories. It's genuinely useful for Devon or Edinburgh clients.

Is the free trial long enough to evaluate

The 7-day free trial on the Pro plan is enough to run one full Site Audit, set up Position Tracking, and test Keyword Magic. It's not enough to see ranking impact.

Can Semrush replace Google Search Console

No. GSC is still the ground truth for your own site's performance. Semrush is a competitor-aware layer on top. Use both.

What's the typical learning curve

Three or four hours of focused use to find the five features you'll actually need. Many freelancers over-complicate the first week by exploring everything.

I funded each account with £500 in late June 2026 and ran the same baseline trades monthly. Buy £100 of a US tech stock, buy £100 of a UK FTSE 100 holding, hold both for 30 days, sell both, convert any USD back to GBP. Same stocks, same dates, same instruction order, documented. The surprise was less about who had the lowest headline fees and more about which one actually behaved sensibly when you wanted to put money into an ISA. One of these does something no other UK broker currently does, and it changes who it suits. Let's go through them.

Robinhood UK Changed the Fractional Game, and the FX Fee Matters

Robinhood UK's marketing pitches "commission-free fractional investing", and in testing, that's genuinely accurate for US stocks. A £100 buy of Nvidia executed at the mid-market USD price with zero platform commission and zero withdrawal fee if you later cash out. The FX conversion built into the trade carried a 0.03 percent spread, which on £100 is 3 pence. For a UK retail investor who mostly wants to buy slices of US tech, this is the lowest total cost of ownership of the three apps tested, and it's not close. The order book routing produced fills within 2 seconds of placing on average, which matched the stated behaviour.

The caveat as of September 2026 is that Robinhood UK is still not offering a stocks and shares ISA, meaning your gains are taxable above the £3,000 annual CGT allowance. If you're trading in sub-ISA amounts anyway, which most new UK investors are, this doesn't matter yet. If you're building a portfolio you'll hold for ten years, it eventually will. The app itself is clean, the interface feels faster than eToro's and more polished than Trading 212's, and the 24/5 trading hours on US stocks are genuinely useful for anyone wanting to trade earnings reports in UK evenings. The fractional allocation went down to 1p worth of stock cleanly, with no "minimum fractional order" friction that some competitors still enforce. For pure execution speed, cost and app quality, Robinhood UK is the strongest of the three in 2026.

eToro: Social Trading That Still Costs Too Much on Small Trades

eToro's pitch is "copy top traders", and in testing, the social feature works as advertised. You can genuinely copy a trader's full portfolio with a one-click allocation. The question is whether that's a good idea, and whether the fees justify the convenience. The platform charges no commission on stock trades, which sounds great, but the 1 percent spread baked into every open-and-close pair on non-US stocks adds up fast. A £100 UK stock round-trip costs you roughly £1.50 in effective spread on top of a £5 withdrawal fee. For small-ticket investors, that's 2 to 3 percent of your capital gone before you've made a trading decision.

Where eToro genuinely helps is diversification for complete beginners. The CopyTrader feature lets you allocate as little as £200 to mirror a specific user's trades, and the risk score system actually highlights the aggressive traders you probably shouldn't copy. Over the test period, a conservative-rated trader I allocated £200 to returned 3.2 percent before fees and 1.1 percent after, which is honestly middling. The ISA situation at eToro is also incomplete. The ISA product exists on paper but is a feature of their broader portfolio service, not full flexibility across the trading app. If you want a UK stocks and shares ISA with real trading flexibility, eToro is not it.

Trading 212: The ISA Winner That Still Feels a Bit Rough Around the Edges

Trading 212 is the one UK newcomers hear about most, and the main reason is simple. It's the only one of the three that offers a free, fully functional stocks and shares ISA, with commission-free trading across US and UK stocks, fractional shares down to 1p, and no account management fees. For a buy-and-hold UK investor wanting tax efficiency on anything above the £3,000 CGT threshold, this is a serious advantage. I funded the ISA account alongside the general account and the ISA worked identically on execution speed, fractional sizing and withdrawal flow.

The caveat is that Trading 212's app is the least polished of the three. The UI occasionally shows loading states that don't resolve for 20 seconds. Chart tools are basic. Order placement defaults to market orders without a prominent limit toggle, which caught me out once on a volatile open. Customer service is also the weakest, with a 48-hour email response window during the test period versus Robinhood's 8-hour average. The 0.15 percent FX fee on USD conversions is slightly higher than Robinhood's 0.03 percent, which on frequent trading of US stocks compounds. For long-term ISA holding, Trading 212 is the sensible pick. For active trading of US stocks, it's not.

What Each One Actually Costs on the Same £1,000 of Activity

Across the test period I logged every fee charged on roughly £1,000 of monthly trading activity split across US and UK stocks. Robinhood UK charged £0.47 in effective FX spread, £0 in commissions, £0 in withdrawal fees. eToro charged £8.20 in effective spread, £5 withdrawal, £0 commission. Trading 212 charged £2.10 in FX spread, £0 commission, £0 withdrawal. For an active trader doing 20 or 30 round trips a month on US stocks, Robinhood is £5 to £8 cheaper monthly than Trading 212 and £15 to £20 cheaper than eToro. For a passive ISA investor, Trading 212's tax-wrapped position wipes out the small FX gap over any time horizon longer than two years.

Deposit Methods, Withdrawal Speed and Cash Handling

Robinhood UK supports bank transfer, debit card and open banking deposits. Open banking deposits cleared instantly. Debit card deposits cleared within 10 minutes with no fee up to £5,000 monthly. Withdrawals back to the funding bank took 1 to 2 working days, consistently. eToro accepts the same methods but adds PayPal and Skrill, with a £5 withdrawal fee on each. Trading 212 supports bank transfer and card, with instant top-ups and free withdrawals. All three held uninvested cash in an interest-bearing arrangement during the test, with Trading 212 offering the highest headline interest at 4.6 percent on GBP balances, Robinhood offering 4.1 percent, and eToro offering a lower rate on balances above a threshold.

Side by Side

Do's and Don'ts

Frequently Asked Questions

Is Robinhood UK regulated by the FCA

Yes, Robinhood UK is authorised and regulated by the Financial Conduct Authority. Client funds are protected under FSCS up to £85,000.

Can I buy fractional shares of UK companies on any of these apps

Yes on Trading 212 and Robinhood UK. eToro supports UK fractional on some stocks but not all, so check per-ticker before relying on it.

Does Robinhood UK offer a stocks and shares ISA yet

Not as of autumn 2026. If an ISA matters for your strategy, Trading 212 is the clear pick for now.

What happens if a broker goes bust

UK brokers are covered by FSCS to £85,000 per customer. Client assets are also held in segregated accounts separate from the broker's own balance sheet.

How does FX spread actually work on US trades

When you buy a US stock in GBP, the broker converts to USD at a rate slightly worse than mid-market. The gap between the mid-market rate and your rate is the spread, usually quoted as a percentage.

Which app is best for an absolute beginner in the UK

Trading 212 for the ISA access and intuitive onboarding, or Robinhood UK for a cleaner app if ISA access isn't an immediate concern.

Can I move my investments between these brokers

Partial in-specie transfers are possible on Trading 212 and eToro ISAs. Robinhood UK supports in-specie transfers in, but outbound transfers are usually processed as sales.

Are there any UK withholding tax quirks on US dividends

Yes. A 15 percent US withholding tax applies to dividends via the W-8BEN form, which all three apps auto-complete. Your ISA doesn't exempt you from this US-side tax, only UK-side tax.