Outside on the road, the familiar van was parked. Since I do love to post an amusing photo on FaceBook, no matter how small the audience that will “get” the joke, I snapped a picture of it and posted it with a caption of “Well that’s telephony and internet in Westerham f*cked then…”
I posted the photo to FaceBook at exactly 09:15. This time is important – there will be a test later.
I then drove to work, thinking no more about it. I have quite a long drive to do; around an hour usually. I work at an ISP called Andrews & Arnold. We are a small ISP with thousands of broadband customers around the UK. Declaring this interest will perhaps explain why I experience fear and dread whenever I see a Kelly van, or an OpenReach van, or any other contractor for BT.
Put simply – and it isn’t just me that thinks this – Kelly have a dreadful reputation in the industry. In our support department, scarcely a week goes by without several cases of lines mysteriously going down, only to later discover that an OpenReach engineer or a contractor to BT (such as Kelly) has “borrowed” our customer’s copper pair (the wires that link the customer’s master phone socket back to the exchange). This borrowing activity obviously kills our customer’s line – both for PSTN and ADSL/FTTC. It is then our problem to get a fault logged and an appointment booked. It is inconvenient and loss making for us to deal with. It is inconvenient for the customer.
Nobody wins. As an ISP, we are fairly well known for getting problems sorted. But these are problems that should never occur in the first place.
Let me run that situation by you again, because I guess to someone not in the industry, it probably sounds far fetched and crazy. But it is a very common (mal)practice :
- An engineer comes out to install a new phone line
- S/he needs to find a “spare pair” (of copper) to run the line back to the exchange
- S/he cannot easily identify which of many is a spare
SO either s/he :
- Doesn’t care, wants to get the job done and closed off, and just steals one
- Does improper tests, uses one that looks unused, but which in fact is still live
Why should we be picking up the support burden of this activity? It is a disgrace.
Regretably it is nearly always circumstancial; our customer vaguely recalls seeing a van and an engineer working near the cabinet, or we hear from the engineer that goes out to fix the broken service that it looks like the pair had been stolen… But by then it is too late to do anything about it; sometimes it is days later.
It was a text telling me my broadband line was down, and timed at 09:26 and 53 seconds. Uh oh. So my semi-joking (semi-serious) prediction on FaceBook was probably right. Just to remind, I posted my photo at 09:15 and just under 12 minutes later, my line went down.
Thinking that this was far too much of a co-incidence, I used our internal systems which interface directly with BT to do a copper line check from the exchange. Yep, sure enough, it was reporting a copper line fault; specifically a “Dis(connect) In Exchange”. So we did a fault report with OpenReach. OpenReach is the division of BT that we buy our lines from. This isn’t actually a terrible process and mostly works quite well. But it isn’t always fast, as an engineer appointment is usually required.
So I wondered if there might be some recourse directly with Kelly. It is important to note that we have no contract directly with Kelly. We simply buy our line from OpenReach, and any faults with it contractually are for OpenReach to sort out (even if they farm some of their work out to contractors such as Kelly).
But here was a pretty cut-and-dry example for assigning blame.
So I re-examined my photo, and realised that luckily I had not just captured the Kelly telephone number, but also the registration number of the van in question. I called the number and reported the problem; that an engineer had been working in my block of flats, that my line had been broken, almost certainly as a result of the engineer mis-using a copper pair being used for my line, and I wanted the engineer to undo what he had done. I was told I’d be called back later in the day. I don’t think the call was treated particularly seriously. And sure enough an hour or more later, I’d heard nothing.
I then had a sneakier idea. I called them again, but this time said the van was causing a problem in the street, and gave the registration number and said I needed to speak to the driver.
This got me through to the fleet manager, who, (rightly enough) unwilling to give me the mobile number of the engineer, was willing to call the engineer whilst I stayed on the line. After confessing that actually it wasn’t a road traffic problem but a technical one, and explaining my ISP affiliation (and saying that this was a common occurrence for us) he rang the engineer.
I only heard one side of the conversation. It sounded like the engineer knew what he had done, and seemed to be conceding that he’d need to go back and fix it! Progress!
The fleet manager was interested in my comments about how Kelly were perceived and said that he’d be interested in an email outlining the problems for onward discussion with other high level managers. He seemed unaware of Kelly’s reputation.
He also mentioned his broadband wasn’t very reliable, so I sent the following :
Hi there *******, Many thanks for talking with me earlier. I hope the engineer (I think his name was **** *****) manages to go back and fix my broken line later today; thanks again for making the appropriate call. It sounds like this may be a handy connection to have made for other reasons though. We (Andrews & Arnold) frequently have issues with work Kelly have done and genuinely without wishing to insult I do feel that Kelly have a bad name in the industry for reasons similar to the ones outlined. We have many examples where Kelly engineers have used copper pairs belonging to our customers to install lines for other end users; totally breaking our service. This has been confirmed to us by either our End User (reporting what BTOR say when the fault is fixed) or by engineers directly. It seems to me that it is quicker and easier to "create a fault" that in all probability another engineer, possibly from another company will need to fix, than to actually install another cable. And this is the nub of the problem. Pressed engineers will take shortcuts, especially when there are no real consequences, and they can close off a job nice and quickly. I'm really not exaggerating when I say we have dozens and dozens of examples of this going on. It is a problem with OpenReach as well. But it's more noticeable and more pronounced with Kelly. I attach a "FaceBook chronology" of this happening at my address in ***** Court, ***** Road, where your engineer was installing a line at Flat *, and in doing so, I assume, broke my line in Flat *. I partly attach this for your amusement, but also to demonstrate that I foretold the problem occurring at 09:15 and, sure enough, at 09:26:53 (approx 11 mins later) my foretelling was proven correct. Of course by then I was driving to work, and only learned of the outage when I arrived in the office and checked my texts. But I hope this (silly) example demonstrates my strong faith that Kelly will break phone lines! I predicted it would happen. And, of course, it did! Partly why we (Andrews & Arnold) are known for reliability is our dogged pursuit of things like this. Had the same information been provided us by a customer we'd have acted on it in the same way. If you are interested in an Internet connection where the people supporting it actually care enough to chase down those responsible for faults and get them fixed, please do have a look at our website www.aa.net.uk. As you can see from the text message, we monitor all our customer lines very pro-actively. Indeed, we send an echo down every single line, every second. ... We are a little more expensive than some, but you get what you pay for. Finally, if there is anything that can be done - in terms of us working with Kellys (despite the fact we have no direct contractual relationship) to help things improve, we'd be pleased to do so. Best regards, and many thanks again. Alex.
Not impolite, I hope you will agree. And the fact that he was interested in the problem was at least a bit impressive. I realise I called them Kellys (they are just “Kelly” singular) a few times. Oh well.
Later in the day I had a call letting me know that Kelly would come back and fix it later that day. This did seem a fairly clear admission of culpability to me. Still later in the day, I had a call from the (different) engineer who’d come and fix the first engineer’s mess. He said on the phone, which I have a call recording of, that the first engineer had indeed used my pair to install the new line for the upstairs flat.
So what I predicted would happen when I snapped the photo at 09:15 had, pretty much precisely, come true. The only difference between this and almost every other time was that this time I had enough evidence to force them to come back and put things right.
I was initially puzzled why access to my property would be needed to fix the fault. Of course, when the first engineer had disconnected my master socket, he hadn’t made it clear which wires in the box were mine. So the second engineer would need to plug a tone generator into my master socket to trace the pair in the box, and reconnect them. Once this was done, the line reconnected and all was well again.
At this point I should mention that I do not just have one ADSL line. I have two bonded, and only one was affected. I also have 3G backup available. Anybody who is serious and demands reliable Internet should have this type of redundancy in place. It is what we (at A&A) are specialists at providing.
So this might have been the end of the story. Not quite though. I asked the second engineer what had happened, and how my line had ended up being the sacrificial lamb. What he said had a worrying and surprising side to it.
The original engineer had apparently not quite followed procedure. Procedure being :
- croc clip the line
- listen for dialtone
- do the “123 test”
- do 17070 to establish the number
- check the number with “Numbering” to ensure it is a stopped line
So my understanding of step 3 of that process is to dial 123, which is a chargeable call, and chargeable calls are denied on stopped lines. But chargeable calls are also deliberately denied on our (A&A) PSTN lines, because our lines are primarily for broadband. So that is a stupid test. 17070 is an engineer’s test facility which *will* work on a stopped line, at least for a while. 17070 will read out the line’s number, and offer a couple of other tests like quiet line and ringback. The numbering section should be able to confirm the line is stopped. In this instance, apparently, the engineer hadn’t done the last step, so had wrongly assumed the line was stopped.
On my way to work today I realised that if the “123 test” really was just dialling 123 on an unknown line then that means making a chargeable call on an unsuspecting and innocent person’s line. Or, if going through a large junction box, the lines of *many* innocent people. That is to say, causing a 31 pence charge for the speaking clock on every single line tested. This is totally unacceptable and very probably illegal. The Communications Act 2003 certainly seems to think it might be : http://www.legislation.gov.uk/ukpga/2003/21/section/125 then there is possible fraud? I am not a lawyer, but this has to be really quite wrong. I know 31 pence is probably not going to kill anyone, but still…
Because I had to be sure that this was really what it sounded like I called the engineer who had fixed my line last night again this morning, and asked him to elaborate on what the “123 test” was. He confirmed it to be as I suspected. He made it sound very much as though this was common procedure; not something that he alone (and the first engineer) did. And I have no reason to doubt that.
Would you notice an incorrect 31p speaking clock charge on your phone bill?
Later in the day today, I spoke to the HR and Quality Manager at Kelly, who had emailed me yesterday to see if we might discuss the situation. He assured me that it definitely wasn’t in the procedures for installing lines to steal copper pairs and it also wasn’t in the procedures to use the “123 test”.
He claimed he had never heard of the test. Overall I wasn’t very impressed with the call. In one sense I am slightly impressed that he bothered to take the time, but on the other hand, I think he wants to tell himself that my situation is a “one off” and not common place. He seemed fairly unwilling to accept – and keen to minimise – my assertion that we see many repeats of these problems in our line of business.
He did commit to investigate fully, and report back to me, though I don’t hold out much hope of ever hearing truthfully if the “123 test” has become an informal norm of a Kelly engineer’s life. I say I don’t hold out much hope simply because to admit that it had become one would be to also admit the theft of 31 pence from potentially thousands of unsuspecting individuals and companies! So, no, I don’t expect that.
I would be delighted if any BTOR or Kelly engineers want to post their views in the comments below. I should say, aside from the first engineer (who was meant to be installing a line for someone else, and not breaking mine) I was impressed by the second engineer, and I was impressed by the Fleet Manager, who took my comments seriously.
My opinion of Kelly hasn’t really changed. I reiterate that we (A&A) are happy to host a meeting for Kelly/OR managers to give real world examples, and lots of them, of this happening. We really wish it didn’t happen, and anything that can be done to reduce incidents has got to be a good thing.
I will finish with my two loss and latency / usage graphs. One for my Kellybroken line, and the other for my untouched line. Yes that is over 10 hours of outage on one line, and happily, no loss outage the other.