DEV Community

Dialphone Limited
Dialphone Limited

Posted on

I Tested 9 UK VoIP Mobile Apps for Notification Reliability. Only 3 Passed.

From my experience the single most common complaint about business VoIP, across every UK provider I have worked with, is: "the mobile app does not ring reliably." So I set up a controlled test last month. 9 providers, 200 calls to each, same hardware, same network conditions.

The methodology

Two test phones: iPhone 15 Pro (iOS 17.4) and Samsung Galaxy S24 (Android 14). Both on the same 5 GHz WiFi, both with notifications enabled and battery optimisation off. Each provider's app was installed, logged in, and left idle.

200 calls were placed to each test number:

  • 100 over 3 minutes spacing (gives the OS no reason to sleep the app)
  • 50 over 15 minutes spacing (realistic business use)
  • 50 after leaving the phone idle for 4 hours (the "overnight" test)

Pass criteria: the phone rings within 4 seconds, every time. "Rang eventually after I opened the app" is a fail.

The results

I will not name the 6 that failed because this is meant to be educational not vengeful, but the scores are real.

Provider ranking (combined iOS + Android pass rate)

Rank Pass rate (iOS) Pass rate (Android) Overall
1 99.5% 97.0% 98.3%
2 98.0% 94.5% 96.3%
3 96.0% 92.5% 94.3%
4 92.0% 73.0% 82.5% FAIL
5 89.5% 68.0% 78.8% FAIL
6 87.0% 62.5% 74.8% FAIL
7 81.0% 45.0% 63.0% FAIL
8 72.5% 38.0% 55.3% FAIL
9 58.0% 21.0% 39.5% FAIL

Honestly not gonna lie, I expected the spread to be maybe 10 percentage points. It was 59 percentage points. The worst provider on the list had 79% of Android calls not ring within 4 seconds.

Why Android scores are so much worse than iOS

Two reasons.

First: push notification delivery on Android depends on Firebase Cloud Messaging (FCM). If the app is in Doze mode (Android battery saver) and the provider does not request a "high priority" FCM message, the notification can delay up to 15 minutes. Most providers do not properly configure high-priority FCM.

Second: Samsung and Xiaomi specifically implement aggressive background-app killing on top of stock Android. If the app has been backgrounded for 2+ hours, Samsung's OneUI often kills it outright. The app must be whitelisted in battery optimisation settings, which is a setting buried 4 taps deep that users never find.

The top 3 providers had one thing in common: they either used CallKit-equivalent (iOS VoIP API) + PushKit plus proper high-priority FCM on Android, OR they ran a persistent foreground service on Android (more battery use, but 99%+ reliability).

The test you can run in 5 minutes

Pick the provider you are evaluating. Install the app. Do NOT tweak battery optimisation settings — use the default installation. Leave the phone idle for 4 hours (overnight). Ask someone to call you 5 times spaced 10 minutes apart.

Count how many rings arrived within 4 seconds.

If the answer is less than 5/5, check again next day. If it is less than 9/10 across 2 days, the app will fail for your staff. Do not buy.

What to ask a provider before signing

  1. Does your Android app use high-priority FCM? (If they do not know, it does not)
  2. Does your Android app run a foreground service option for users who need 100% reliability?
  3. Does your iOS app use PushKit and CallKit? (Industry standard, should be yes)
  4. What is your mobile app measured notification success rate over the last 90 days? (If they cannot answer with a number, they are not measuring)

My test had 1 provider who flat out did not know what FCM was. They were selling to UK SMBs. Not great.

Where DialPhone came out

I will be honest because it is my employer: we came out 2nd. 96.3% combined. We lost points on Samsung Galaxy S24 background killing, which we are fixing with a foreground-service opt-in in our April release. The winner was a provider I respect and will probably benchmark against again in Q3.

The only thing that matters for you as a buyer: test the app yourself, on the phones your staff actually use, before you commit.

Top comments (0)