MFR

The UKmix topic indeed has the exact panel sales but the multiplied up 'market' figures remain estimates or interpretations of what was sold to the public.

The reasons for not including sales from before 1994 are 1) These figures are estimates derived from sample sales and are not intended to be 100% accurate, just a best guess for public consumption, although they may be fairly close in many cases. Until 2013 certifications were based on copies shipped by record companies to the UK retail market. Such figures were known accurately to the record company making the application. However, even from 1994 onwards retail sales figures were still estimates, but increasingly accurate. In reality what is being certified today is streams with an a reasonably accurate estimate of prior physical sales and download sales from 1994 included. 2) Even if a decision were made to include approximations from prior to 1994 the OCC would have to have estimates for everything ever released even if the single had only a few thousand sales prior to 1994, not just the big hits because so many songs are available to be streamed. Their database only debited in 1994 so that would be an impossibility.

This shouldn't be seen as surprising because the 1989 silver certification has no bearing on new certifications in the way that some might think. Current certifications are based on sales and sales equivalents since 1994 only. In effect everything released prior to this was on zero sales at that point. This was not an issue while certification was still up to record companies using their own data. In this case the 1989 silver certifications blocks any subsequent silver certification on reaching 200,000 sales equivalents since 1994 and the next activity on this single would be gold on achieving 400,000 equivalent sales since 1994. Unless you were expecting to already be on 400,000 since 1994?

The answer is that until 2013, when automated certifications started, record companies could decide whether they wanted to apply to the BPI for certification, and in 1990 CBS UK (as Sony was then ) chose not to apply in case. I would guess it was because Maria McKee was not actually signed to them.

I don't think I have the Music Week with that article any more. It was either December 9th or 16th 1995. I can't remember the name of the WEA person quoted either, but he was probably the executive on the WEA dance music imprint that the Outhere Brothers were on. The figures given were 560,000 for Don't Stop and 525,000 for Boom Boom Boom. I may be wrong about the executive's exact words about Boom Boom Boom, but it was something very much like "although it didn't sell quite as many as its predecessor many people, including myself, preferred it". So there was no doubt that Music Week had got the two figures the right way round. As I said the retail total for 1995 was 105% of the deliveries. Between 1997 and 2001 the highest figure was about 92% in 1998, when the chart year was 6 days longer than the deliveries year. All the other years were below 90%.

In hindsight, it seems that it was a mistake for OCC and Millward Brown to increase the multipliers in 1994, when for singles it was increased from 15, which Gallup had used from the start of 1992 onwards and which OCC began with, to 17. This occurred in August following a survey by Millward Brown, but the whole year was re-cast at the higher multipliers, It looked OK in late 1994 and early 1995 but fell away mysteriously from quite early (March? April?) in 1995. From this point on I would think most singles would appear to have sold more at retail than the number of copies shipped. This phenomenon was already there in 1994-1995 (Think Twice only certified single platinum, but 'selling 1,235,000 at retail). But it was the 2 Outhere Brothers number 1s that highlighted it, because an article in Music Week in December 1995 about dance music successes in the singles chart listed shipment figures supplied by the record company people interviewed. Both those Outhere Brothers singles had retail estimates above 600,000, but a WEA person gave their shipped figures as below 600,000 in both cases, which is why they only got certified gold. Worse, at retail it was the second hit that sold more by a reasonable margin, but the first one shipped more by a reasonable margin. It seemed to show the multiplier was too high early in the year and way too high later on. Amongst the sales given, if I remember correctly,, Here Comes Th Hotstepper by Ini Kamoze, from the very start of the year and released in late 1994, did ship more than retail sold, and for Don't Give Me Your Life by Alex Party the two figures were about level, but the other later ones had retail estimates above the shipped figures. In fact the total number of singles sold at retail in 1995 was 5% higher than the 'trade deliveries' figure for 1995 that was compiled by the BPI. Of course, a single sold in 1995 might have been shipped in 1994, but the same applies at the other end of the year. During 1997 to 2001, the total sales in DUS terms were in the 85% to 90% range compared to deliveries in each year, except 1998, when it was about 92% I think. 1998 had a 53rd sales week at retail, which can partly account for the higher percentage that year. So in mid-1995 when supermarkets were about to be added to the defined universe, the multiplier was 17. At the start of October when they were added - presumably they test-sampled for a while before officially adding them to the chart sample - the multiplier was reduced to 16, and then became 15 from the start of 1996. That seems telling. Nevertheless, they later reinstated 17 for the whole of 1995 when quoting figures, although Hit Music had used 16 in its year-end data at the time. The 99% point was reached at the start of the fourth quarter of 1996, so only 6 months before DUS reporting began. The number of shops in the defined universe jumped from under 3,500 to just over 3,900 that week. Yes, certifications were based on trust. It was mostly the same dozen or so executives that would have been signing off the majority of the awards anyway in the letters or emails to the BPI, so there was probably little reason to disbelieve. Occasionally an award looked wrong or dubious, but certifications were seen as a good promotional tool in those days. If anyone had clearly been abusing the BPI's trust I expect they would have looked into it, but I'd say that by and large everyone played by the rules. You can't take any award and use it to prove anything Alan Jones used to say. But look at whole years and what individual record companies did and patterns emerge. Sony and BMG were the worst certifiers, but had good periods and less good periods. Virgin, EMI, A&M, London, indeed most of Polygram as was, and WEA were good to excellent. Some record companies sometimes seemed to ignore silver for singles, as if perhaps not important enough (Sony particularly), or didn't certify singles silver if the album was certified already (Sony 1987-1992). Perhaps not surprisingly companies that tended to wait until they had several records to certify missed the odd one or two - maybe this was Sony and BMG's true failing in this area, rather than indifference. Years ago I compared certifications with retail estimates for the early DUS years. Late summer and then the Christmas rush followed perhaps by time off at home for key executives were perhaps the times of year that apparently qualified singles were least likely to get certified. If a single could get to 250,000 - the old level for silver - its chances seem to go up a notch. Maybe this was because they were above 200,000 for longer while still active rather than because someone was stuck in the past. Strangely, singles above 375,000 also suffered. Waiting for a gold certification that was never reached may have been the reason. In terms of years, before auto-certification there were many silver singles from 1994 to 1996 above 400,000 and gold singles above 600,000. Sometimes quite considerably. Over the course of a year the same pattern could emerge. For 1994 to 1996, and also with 1990 and 1991 the appearance is of the multiplier being too high. At one point I thought that rather than decide on a multiplier for a particular year and a different one for the next year, it might have been better to stick with one for a longer period as Gallup had. Eventually, I decided that perhaps the question to ask for 1994 to 1996 is "what would the DUS figure have been if the 99% point had already been reached?". When DUS reporting began the effective multiplier, as it became obsolete was around 15.7, so I would suggest somewhere around that figure would be reasonable for 1994 to 1996 on the basis that 15 was starting to look a little low when Gallup's contract ended. But it would vary from record to record, or certainly from genre to genre.

I think there's a lot of misunderstanding about this. OCC will have historical records about how many shops were in the sample each week from when they took over from Gallup in 1994. Gallup historical data probably not. The difference between the sales total from 1994 to 1996 under the two methods is this - 1. In the panel sales method sales were reported to the industry on the basis of the numbers sold in a permanent 250 shop universe. The actual number of shops reporting each week was irrelevant. It's just that it changed from week to week and so to have a consistent picture sales reports were based on a standard of 250 shops and these were the 'panel sales'. People outside the industry (us, for example) wanted to know what the total sale was, not just what a standard panel sold. So they multiplied the panel sales by a figure called the multiplier which was reckoned to give an answer approximately equal to the total sold in the whole market. But it was approximate. Some shops not sampled didn't sell as many as similar shops that were on the panel. But you don't know that if you haven't sampled the data. There was a lot of '10 of these shops sold x copies in our sample, so another 10 shops not in the sample would also sell x of the same record'. This can work if all 20 shops were part of the same chain and of similar size, and maybe in the same geographical area. But then there were independent shops. They were more likely to be individual and a shop like this not on the panel was less likely to be a direct match for another independent shop that was. Also independent shops that were on a register of shops, used to determine the market size, might have gone out of business without the chart compilers knowing, if they didn't used it. Then there were supermarkets, of which none were on the chart panel until 1995, so how to estimate what they had sold? Would 10 supermarkets sell the same as 10 Woolworths, or 10 HMVs, or 10 independent shops? Something would have been sold. but without having any in the sample, how could the compilers be sure how many? Mail order companies, motorway service stations and market stalls were types of music retailers that were never in the sample. But all could be taken account by sampling 1,000 shops that were in major chains plus some indies and multiplying the figures up. Usually by too much. 2. The DUS method works differently. If you sampled half the large HMVs you could probably double the sale you got to come up with a figure for both halves. But it told them nothing about Woolworths sales. Of course hey sampled data from some Woolworths for that. And other chains similarly. These were all 'defined' because they occurred somewhere in the sample, hence 'Defined Universe Sales'. Independent shops were either in or out. The ones that were not in were therefore assumed to have sold nothing because they couldn't be matched to something that was in. The same was true for mail order companies, motorway service stations, market stalls etc. Record companies knew what they had sold to such places and could adjust the DUS upwards to take account of the percentage of stock sold to these places, if they wanted to. But the key thing with DUS is they let the indutrsy know what was 'defined' and reported only what was estimated to have been sold in that. So the figures in 1 were higher than in 2 because 1 was an estimate of sales in all shops, but 2 was an estimate only for the parts you could estimate accurately for, and leaving out the totally unknown parts that might not be trading now anyway. At least in a digital age, and with streaming, everything is in, unless it's not known, in which case it's still out. But the percentage covered must be virtually 100% now.

This is the top 100 of that chart, with the year and year-end position also indicated. It's the chart that Hit Music didn't quite do - based on the panel sales for 1990 to 1996, and using the multipliers that the chart compilers settled on, rightly or wrongly, for each year - and then DUS from 1997. So at least for 1994 to 1996 the sales are higher than OCC lists often allow for. I can respond to queries about the rest of the chart, although my time for chart chat is quite restricted these days. Looking at the Bruno Brookes list, they seem to have missed The Millennium Prayer. One plus, on the other hand, is they seem to have remembered that The Righteous Brothers version of Unchained Melody was still selling in early 1991. Many lists cut it off at the end of 1990. 1 1997 1 CANDLE IN THE WIND '97 / SOMETHING IN THE WAY YOU LOOK TONIGHT Elton John 2 1995 1 UNCHAINED MELODY / WHITE CLIFFS OF DOVER Robson & Jerome 3 1994 1 LOVE IS ALL AROUND Wet Wet Wet 4 1997 2 BARBIE GIRL Aqua 5 1998 1 BELIEVE Cher 6 1997 4 PERFECT DAY Various Artists 7 1991 1 (EVERYTHING I DO) I DO IT FOR YOU Bryan Adams 8 1999 1 BABY ONE MORE TIME Britney Spears 9 1997 3 I'LL BE MISSING YOU Puff Daddy / Faith Evans 10 1992 1 I WILL ALWAYS LOVE YOU Whitney Houston 11 1996 1 KILLING ME SOFTLY Fugees 12 1996 2 WANNABE Spice Girls 13 1998 2 MY HEART WILL GO ON Celine Dion 14 1995 2 GANGSTA'S PARADISE Coolio 15 1997 13 NEVER EVER All Saints 16 1995 5 THINK TWICE Celine Dion 17 1998 12 HEARTBEAT / TRAGEDY Steps 18 1997 5 TELETUBBIES SAY EH-OH Teletubbies 19 1995 3 I BELIEVE I UP ON THE ROOF Robson & Jerome 20 1996 3 SPACEMAN Babylon Zoo 21 1998 3 IT'S LIKE THAT Run DMC 22 1994 2 SATURDAY NIGHT Whigfield 23 1995 6 EARTH SONG Michael Jackson 24 1996 10 2 BECOME 1 Spice Girls 25 1998 4 NO MATTER WHAT Boyzone 26 1995 10 WONDERWALL Oasis 27 1991 2 BOHEMIAN RHAPSODY / THESE ARE THE DAYS OF OUR LIVES Queen 28 1997 8 TORN Natalie Imbruglia 29 1995 4 BACK FOR GOOD Take That 30 1999 2 BLUE (DA BA DEE) Eiffel 65 31 1996 4 SAY YOU'LL BE THERE Spice Girls 32 1994 3 STAY ANOTHER DAY East 17 33 1995 9 MISSING Everything But The Girl 34 1997 6 MEN IN BLACK Will Smith 35 1990 1 UNCHAINED MELODY Righteous Brothers 36 1999 3 THE MILLENNIUM PRAYER Cliff Richard 37 1998 7 CHOCOLATE SALTY BALLS Chef 38 1998 5 C'EST LA VIE B*Witched 39 1999 4 MAMBO NO. 5 Lou Bega 40 1998 25 ANGELS Robbie Williams 41 1996 5 RETURN OF THE MACK Mark Morrison 42 1997 7 DON'T SPEAK No Doubt 43 1998 8 GOODBYE Spice Girls 44 1995 13 FATHER AND SON Boyzone 45 1997 10 SPICE UP YOUR LIFE Spice Girls 46 1999 5 9PM TIL I COME ATB 47 1996 6 OOH AAH...JUST A LITTLE BIT Gina G 48 1996 7 THREE LIONS Baddiel / Skinner / Lightning Seeds 49 1995 7 FAIRGROUND Simply Red 50 1997 9 TUBTHUMPING Chumbawamba 51 1996 13 UN-BREAK MY HEART Toni Braxton 52 1999 6 LIVIN' LA VIDA LOCA Ricky Martin 53 1996 8 CHILDREN Robert Miles 54 1995 16 COTTON EYE JOE Rednex 55 1999 7 THAT DON'T IMPRESS ME MUCH Shania Twain 56 1996 11 DON'T LOOK BACK IN ANGER Oasis 57 1993 1 I'D DO ANYTHING FOR LOVE (BUT I WON'T DO THAT) Meat Loaf 58 1996 9 MYSTERIOUS GIRL Peter Andre 59 1994 4 BABY COME BACK Pato Banton 60 1998 32 TOGETHER AGAIN Janet Jackson 61 1996 14 BREATHE Prodigy 62 1995 8 YOU ARE NOT ALONE Michael Jackson 63 1997 11 MMMBOP Hanson 64 1998 6 HOW DO I LIVE LeAnn Rimes 65 1997 12 D'YOU KNOW WHAT I MEAN Oasis 66 1999 8 SWEET LIKE CHOCOLATE Shanks & Bigfoot 67 1995 22 I'LL BE THERE FOR YOU Rembrandts 68 1994 5 I SWEAR All-4-One 69 1998 9 GHETTO SUPASTAR Pras Michel 70 1999 9 FLAT BEAT Mr Oizo 71 1997 14 I BELIEVE I CAN FLY R Kelly 72 1990 2 NOTHING COMPARES 2 U Sinead O'Connor 73 1999 10 WHEN THE GOING GETS TOUGH Boyzone 74 1996 12 HOW DEEP IS YOUR LOVE Take That 75 1990 4 ICE ICE BABY Vanilla Ice 76 1997 15 MAMA I WHO DO YOU THINK YOU ARE Spice Girls 77 1997 22 TOO MUCH Spice Girls 78 1995 11 BOOM BOOM BOOM Outhere Brothers 79 1998 11 MUSIC SOUNDS BETTER WITH YOU Stardust 80 1990 3 SACRIFICE I HEALING HANDS Elton John 81 1998 10 TRULY MADLY DEEPLY Savage Garden 82 1995 12 COUNTRY HOUSE Blur 83 1998 45 WHEN YOU'RE GONE Bryan Adams & Melanie C 84 1998 13 VIVA FOREVER Spice Girls 85 1999 11 BRING IT ALL BACK S Club 7 86 1998 14 THREE LIONS '98 Baddiel I Skinner / Lightning Seeds 87 1995 14 DON'T STOP (WIGGLE WIGGLE) Outhere Brothers 88 1993 2 I CAN'T HELP FALLING IN LOVE UB40 89 1996 15 FIRESTARTER Prodigy 90 1999 12 BETTER OFF ALONE Alice Deejay 91 1993 3 ALL THAT SHE WANTS Ace Of Base 92 1997 16 I WANNA BE THE ONLY ONE Eternal I Be Be Winans 93 1999 13 GENIE IN A BOTTLE Christina Aguilera 94 1994 67 WHATEVER Oasis 95 1999 14 PERFECT MOMENT Martine McCutcheon 96 1998 15 DOCTOR JONES Aqua 97 1992 2 RHYTHM IS A DANCER Snap! 98 1999 15 BOOM BOOM BOOM BOOM Vengaboys 99 1993 6 MR BLOBBY Mr Blobby 100 1998 17 I DON'T WANT TO MISS A THING Aerosmith

If you click in either of the date columns and then on the current month that is displayed (i.e. February 2018 at the moment) you can go back a year at a time to any previous February. You can go left to the previous month in the year you have reached or right to a subsequent month. It doesn't then take long to create a range from, for example, January 1991 to July 1994, and view all awards in that period, or refine the selection still further using the other columns. There may be a better way. I'm not sure how easy it is to build up a list of acts with the most awards, but you can sort your selection by artist name and work out who comes up the most over the whole period, though this is not something I intend to check out.

UK record industry certifications, since the BPI launched the official awards scheme in 1973, can be found on their website. https://www.bpi.co.uk/bpi-awards/

Official BPI awards used to be presented only to the record company, the producer(s) of the album or single in question, the artist and the artist's manager. A few years ago the scheme was widened to allow discs to be presented to anyone the company wished to reward for their contribution to the success of whatever the release was. But - record companies have made unofficial awards of presentation discs to anyone they wished to, based on any criteria they liked, regardless of what the actual total sale was. This has been going on for decades and presumably this practice continues. So, unless it's been missed by OCC and the BPI this could be an unofficial award.

Perhaps somewhat counter-intuitively total singles sales increased in 1984 over 1983 because there were a lot of million plus sellers (Band Aid, 2 x FGTH, Stevie Wonder, George Michael, Last Christmas by Wham! nearly there at year-end). If each of these had sold only half of what they did overall sales would have been down slightly. Indeed average sales at positions lower down the chart were significantly lower than in the late 1970s. In 1985 overall sales did drop, but with only one million-seller perhaps they had otherwise held up OK, for one more year at least.

Apart from Candle In The Wind '97, which is at 9x Platinum.

Yes, by about 15,000 from the 1990s figures, although technically it had been over-estimated by most of that 15,000 originally.

What happened to What Do You Want To Make Those Eyes At Me For by Emile Ford & The Checkmates? Possible spolier alert if this is planned as a series of reviews of old issues of Record Mirror, but the next time Alan visited the subject of UK million-sellers he had decided that this track was a worldwide million-seller, but had not sold a million in the UK.