Category Archives: Writing

Hemingway (logo)

Do your headlines have a Beyoncé-level of engagement?

Like the lead para­graph in a news sto­ry or the­sis in an essay, your head­line is your one true sen­tence: the sin­gle most impor­tant asset for cap­tur­ing atten­tion in the feed.

Hem­ing­way is Sharethrough’s new AI-pow­ered head­line ana­lyz­er, an easy-to-use, pub­licly avail­able tool that puts a wealth of pro­pri­etary data sci­ence and lin­guis­tic analy­sis at your fin­ger­tips for the first time. This new tool is free for any­one look­ing to nav­i­gate the new pres­sures and demands in con­tent mar­ket­ing, help­ing them ana­lyze and quick­ly improve the qual­i­ty of their head­lines, opti­miz­ing for both impres­sion and engage­ment.

Below are the intial results for this arti­cle’s head­line (pre-Bey­on­cé).

Hemingway (actual headline result)

Using this analyser, I was able to push my Head­line Qual­i­ty Score from 62 to 79%. I’m not sure adding Bey­on­cé improved your lev­el of engage­ment after you arrived, but you clicked on the head­line though, did­n’t you? Appar­ent­ly that’s 98% of the prob­lem solved.

How does it work?

The Head­line Qual­i­ty Score is based on a mul­ti­vari­ate lin­guis­tic algo­rithm built on the prin­ci­ples of Behav­ior Mod­el the­o­ry and Sharethrough’s neu­ro­science and adver­tis­ing research. The algo­rithm takes into account more than 300 unique vari­ables, includ­ing EEG data and Nat­ur­al Lan­guage Pro­cess­ing, enabling your native ads to cap­ture atten­tion, increase engage­ment and deliv­er a stronger impres­sion.

Basi­cal­ly, it offers sug­ges­tions to improve click­throughs — but’s it not going to write bet­ter head­lines for you.

Back to work, you Hem­ing­way-wannabe.


Chaos Theory (as a Management Style)

For generations, science allowed us to think we could control nature. Today we know better (thanks to Chaos Theory). How could this make us better managers?

Sir Isaac New­ton’s deter­min­is­tic claim that we can pre­dict future events with absolute cer­tain­ty stood firm for 300 years — then along came Kurt Gödel’s Incom­plete­ness The­o­rem and Wern­er Heisen­berg’s Uncer­tain­ty Prin­ci­ple. Apply­ing this new sci­ence to the most wide­ly used man­age­ment style (i.e. direc­tive) and com­par­ing it to more empow­er­ing tech­niques, I look at how this could inform man­age­ment tac­tics (as well as social media pol­i­cy and cor­po­rate social respon­si­bil­i­ty).

The boring science‑y bit

mathematical formula written in chalk on blackboardIn 1931, Kurt Gödel declared a for­mal proof that every sys­tem (even the all-empass­ing Prin­cip­ia Math­e­mat­i­ca) con­tains incon­sis­ten­cy, and is there­fore incom­plete. In 1927, Heisen­berg’s uncer­tain­ty prin­ci­ple declared that all physics (when exam­ined close­ly enough) con­tains a degree of chaos. This was in con­trast with estab­lished New­ton­ian deter­min­ism that saw the uni­verse as a giant clock — if we could only see the cogs in enough detail, we could pre­dict future move­ments. When com­bined with oth­er the­o­ries (such as ran­dom Brown­ian Motion, Lorenz’ But­ter­fly Effect and Schrödinger’s Quan­tum Mechan­ics), a move­ment devel­oped that came to be known as Chaos The­o­ry. This embraces the idea that we can nev­er tru­ly fore­see an out­come, because small fluc­tu­a­tions can cause large long-term effects.

More recent­ly, in books on macro-eco­nom­ics such as Freako­nom­ics and The Tip­ping Point, there has been an under­stand­ing that growth can­not be infi­nite (e.g. the occu­py move­ment) and an aware­ness of wider sus­tain­abil­i­ty issues. This isn’t a hip­py-dip­py resur­gence of 60’s flower-pow­er — it’s actu­al­ly a nat­ur­al result of look­ing ever more deeply at what was pre­vi­ous­ly only thought of in abstract terms. Rick Levine and Christo­pher Locke dis­cussed the prob­lems of direc­tive man­age­ment styles in their book The Clue­train Man­i­festo. In it, a major study showed that although bark­ing orders at employ­ees often gen­er­at­ed high­er prof­its in the short term, (when com­pared with the long-term gains of more empath­ic man­age­ment tech­niques) it is actu­al­ly unprof­itable in the long-term — the man­age­ment equiv­a­lent of King Canute dar­ing the tide to change.

Today’s management structure prevents information flow

So how do these sci­en­tif­ic and high-lev­el math­e­mat­ics the­o­ries apply to man­age­ment styles — what could they pos­si­bly both have in com­mon? In each case, they lis­tened to the details — instead of ignor­ing them (because they did­n’t fit the estab­lished pat­tern). This often a pre­cur­sor of inno­va­tion — and why small­er com­pa­nies can do this bet­ter than larg­er ones. Chaos The­o­ry demon­strates that, (as a man­ag­er) it’s sci­en­tif­i­cal­ly impos­si­ble to pre­dict what will hap­pen. Direc­tive, short-term man­age­ment pat­terns don’t lis­ten for the details — they deter­mine large-scale changes from pre­vi­ous expe­ri­ence. As glob­al weath­er will tes­ti­fy — what hap­pened yes­ter­day — or last year — isn’t nec­es­sar­i­ly the best indi­ca­tor of what will hap­pen tomor­row.

Inter­nal com­mu­ni­ca­tions with­in mul­ti-lev­el man­age­ment organ­i­sa­tions are not con­struct­ed to allow these details to be fil­tered upwards. In a typ­i­cal man­age­ment meet­ing, there’s only time for the larg­er prob­lems to be dis­cussed, so small­er prob­lems must be ignored — until they grow large enough to be on the agen­da (requir­ing more expen­sive solu­tions). This is also true for many effi­cien­cy and stream­lin­ing process­es — man­agers spend so lit­tle time on the ‘shop floor’ that they are unaware of improve­ments that are sug­gest­ed by those who are clos­est to the prob­lem — the work­ers.

When the going gets tough, the CEOs get out

What appears to be a quick easy fix (such as clo­sures and lay-offs) can show as instant prof­it on this year’s bal­ance sheet — pay­ing for the expen­sive CEO’s gold­en hand­shake, but will typ­i­cal­ly back-fire. In addi­tion, it pro­pos­es a ‘boom and bust’ men­tal­i­ty that caus­es many CEOs to lose their jobs (as soon as the bust hits). High dra­ma makes for great head­lines, but poor man­age­ment. As with cli­mate change, there may be no sin­gle rad­i­cal solu­tion that solves a major prob­lem com­plete­ly — but a large num­ber of small­er improve­ments (when added togeth­er) can pre­vent the need for dra­mat­ic action.

There are oth­er ways to solve this prob­lem more cre­ative­ly — David Cote (Hon­ey­well)Dan Price (Grav­i­ty) and Bob Chap­man (Bar­ry-Wehmiller) per­haps being the most famous exam­ples, but in recent times FedEx, Hewlett-Packard, and The New York Times have all cut base pay (with most low­er­ing man­age­ment salaries more than work­ers) instead of let­ting peo­ple go. Even Lar­ry and Sergey at Google only take a $1 annu­al salary. In Japan, a pop­u­lar belief in busi­ness ethics is that busi­ness­es (and peo­ple) who pur­sue mon­ey first even­tu­al­ly fail — most notably employed by Haru­ka Nishi­mat­su, who humbly wait­ing in line for food with his employ­ees and took the bus to work when times got tough, as good lead­ers should fight along­side their troops. Simon Sinek used a sim­i­lar bat­tle-based anal­o­gy (but the same fam­i­ly motif) as the basis of his book Lead­ers Eat Last.

Since the 1980s, much of busi­ness ide­ol­o­gy has been influ­enced by mil­i­tary tech­niques (e.g. goals, strat­e­gy, objec­tives, tac­tics) — how­ev­er, the com­rade­ship fac­tor has been con­ve­nient­ly left out. This just does­n’t add up.

Social media — “The Truth Will Out”

Tax avoid­ance schemes even­tu­al­ly come home to roost. Get­ting the state to pay for Wal­mart’s employ­ee ben­e­fits (while the com­pa­ny makes record prof­its) is just not sus­tain­able — and the new-found pow­er of con­sumers in social media is the best place to dis­rupt this sort of care­ful­ly-planned (and deter­min­is­tic) mar­ket­ing plan. Social media clos­es the feed­back loop, allow­ing infor­ma­tion to freely bub­ble to the top.

From William J. Conaty, who ran human resources at Gen­er­al Elec­tric (GE) for 14 years:

“Peo­ple have long mem­o­ries. They’ll remem­ber whether they think they were dealt with equi­tably.”

CSR and fair compensation

Man wearing suit and tie stares into camera, as dirt road recedes into the distanceWhat, then is the most sci­en­tif­i­cal­ly accu­rate man­age­ment style that best depicts a mod­el of real­i­ty? How can we take advan­tage of broad advance­ments in sci­ence and math­e­mat­ics to be more effec­tive, bet­ter under­stood, with more cus­tomers, and achieve high­er prof­its (with hap­pi­er and more moti­vat­ed staff)?

Sim­ple — be more humane when man­ag­ing fel­low humans. This is the use of sup­port­ive instead of direc­tive man­age­ment tech­niques. In a busi­ness sense, it leads to more prof­it. For employ­ees, they are hap­pi­er and feel val­ued. Cus­tomers ben­e­fit through a bet­ter lev­el of ser­vice.

This is why Hen­ry Ford dou­bled the min­i­mum wage in 1914, why the Cad­bury broth­ers cre­at­ed the town of Bournville for their staff and pio­neered pen­sions in 1879, and more recent­ly Face­book have cre­at­ed their own com­pa­ny town — these (even­tu­al­ly) lead to high­er prof­its. Fair com­pen­sa­tion (and recog­ni­tion — which is free, after all) is often all that employ­ees ask for. These are some of the ear­li­est exam­ples of Cor­po­rate Social Respon­si­bil­i­ty — which seems these days to be com­plete­ly divorced from employ­ee ben­e­fits, and has turned into a form of cor­po­rate phil­an­thropy (i.e. for those out­side the com­pa­ny) instead.

Typ­i­cal­ly, small­er fam­i­ly-run busi­ness­es sup­port their employ­ees, and lis­ten to cus­tomer and pro­duc­tion’s poten­tial prob­lems — and are thus able to fix them while still in their infan­cy. This long-term approach is often lam­bast­ed by more ‘prof­it-dri­ven’ man­age­ment exec­u­tives — but we should be think­ing in terms of being in sync with our cus­tomers, clients and col­leagues for decades — not try­ing to rip them off as quick­ly as pos­si­ble and hop­ing there’ll be a new suck­er born every minute.

Business relationships are a conversation (not an argument)

By free­ing up the infor­ma­tion flow, respect­ing each oth­er’s prac­ti­cal, man­age­ment, and user expe­ri­ence, we can cre­ate high­ly-opti­mised yet flu­id and respon­sive solu­tions that evolve organ­i­cal­ly over time. By lis­ten­ing to our col­leagues, we learn to embrace chaos — and respond quick­ly to the unknown because we knew it was always there.

In our new knowl­edge econ­o­my, thought­ful appli­ca­tion of new sci­ences and tech­nol­o­gy using the above tech­niques will inevitably lead to brand loy­al­ty, less employ­ee churn, deep­er cus­tomer engage­ment and high­er prof­it mar­gins.

Isn’t that what we all want?

But how?

The best way to go with the (chaot­ic) flow isn’t to throw your hands up in despair — roll up your sleeves, use good judge­ment and demon­strate lead­er­ship. It’s best encap­su­lat­ed by Saint-Exupéry (author of The Lit­tle Prince) in this TED talk by Julia Galef about the ‘Scout Mind­set’:

“If you want to build a ship, don’t drum up your men to col­lect wood and give orders and dis­trib­ute the work. Instead, teach them to yearn for the vast and end­less sea.”

Million Dollar Homepage

IAB: A U‑Turn on the Ad-Blocking Superhighway?

Ad-blocking is the new normal. With the Interactive Advertising Bureau (IAB) having launched its LEAN Ads program worldwide, I look a little closer at the initiative – and what it implies for the future of online advertising.

While I agree with the the­o­ry of the LEAN ini­tia­tive (which stands for light, encrypt­ed, ad-choice sup­port­ed, and non-inva­sive), the imple­men­ta­tion leaves a lit­tle to be desired. Less place­ment oppor­tu­ni­ties for pub­lish­ers and more con­straints for dis­tri­b­u­tion plat­forms seem an Orwellian reac­tion to an indus­try still reel­ing from the arrival of HTML5.

pagefair-mapToday, almost one in five Inter­net users in the UK (and ris­ing) have an ad-block­er installed. Adver­tis­ing rev­enue is being wast­ed on unseen ads, ad fraud and ‘bots, while scripts, videos and bloat­ed band­width are inflat­ing mobile data plans. The tar­get audi­ence and brands are cry­ing ‘foul’. And although LEAN address­es some of these tac­ti­cal con­cerns, it fails to address the broad­er prob­lems.

For­tu­nate­ly, there are ways through this thorny prob­lem but, much like glob­al warm­ing, we’re not going to like it; agen­cies, dis­tri­b­u­tion plat­forms, pub­lish­ers and clients are all going to have to work togeth­er if we’re to move for­ward.

First­ly, as always, we need to get cre­ative. Take the hum­ble 200kb online ad; often del­e­gat­ed to art­work­ing teams, many with mea­gre bud­gets, fast turn­around times and low expec­ta­tions.

Incor­po­rat­ing dig­i­tal inno­va­tion, such as dynamic/rotating con­tent; lever­ag­ing speed using Con­tent Deliv­ery Net­works as well as pro­gram­mat­ic and oth­er user-tar­get­ing tech­niques; and devel­op­ing con­tent-led cre­ative – instead of just con­tain­ers for con­tent deliv­ery – may even­tu­al­ly endear the user to brands and increase engage­ment. This approach will cre­ate ads that evolve and can last an entire cam­paign – simul­ta­ne­ous­ly reduc­ing media spend while increas­ing click­throughs. Block­ing ad-block­ers is a road to nowhere.

Sec­ond­ly, lead­ers in this area (such as Guardian Labs) are invit­ing users to become part of the equa­tion. An exten­sion of the IAB-approved ‘AdChoic­es’ con­cept, Google’s Con­trib­u­tor plat­form for Dou­bleClick (which is yet to roll out to the UK), allows ‘sub­scribers’ to pay a month­ly fee to remove ads. How­ev­er, this will only work if all ads are removed in the sub­scrip­tion, and the prof­it mod­el replaces the rev­enue stream (and doesn’t increase it). If there’s one thing online busi­ness­es should learn, it’s that trans­paren­cy is key to suc­cess.

crystal_page_load_timesLast­ly, pub­lish­ers, clients and media plan­ners seem to have opt­ed for quan­ti­ty, not qual­i­ty. Those that work hard­er with their part­ners, lever­ag­ing brand depth instead of reach, to gain the first-mover advan­tage (reduc­ing impres­sions and incor­po­rat­ing native/sponsored/branded con­tent) will be the first to reap the low-hang­ing fruit; leav­ing com­peti­tors, pay­walls and ad block­ers scram­bling in their wake.

There has been some size­able changes in the dig­i­tal dis­play indus­try in 2015, but for a long time users have always want­ed the same thing from adver­tis­ing: make it use­ful.

Show me what I need, just before I need it. 

Most users don’t want to block all adver­tis­ing; they just want to see adver­tis­ing that is appro­pri­ate to them (by def­i­n­i­tion, ads not intend­ed for them are – at the very least – poor­ly tar­get­ed). We have many more cre­ative dig­i­tal tools to enable this to hap­pen.

The indus­try has alien­at­ed our cus­tomers with irrel­e­vant adver­tise­ments force-fed to them en masse – let’s work hard (and togeth­er) to get them back on board.

They’ll thank us for it.

“Mil­lion Dol­lar Home­page” © 2005 Alex Tew
“Ad Block­ing Usage by Coun­try”© 2015 PageFair/Adobe
“iOS Page Load Time in Sec­onds” © 2015 Mark Wilson/Beta News

Fire & Brimstone at the Buncefield Oil Depot, Hemel Hempstead

0520 GMT: Cal­cu­la­tions show that Tank 912 would be com­plete­ly full and start­ing to over­flow. Con­tin­ued pump­ing caus­es fuel to cas­cade down the side of the tank and through the air, lead­ing to the rapid for­ma­tion of a rich fuel/air mix­ture that col­lects in bund A — the low enclo­sure sur­round­ing 912 and neigh­bour­ing tanks.

0538 to 0546 GMT: CCTV footage shows vapour of escaped fuel start to flow out of the north-west cor­ner of the enclo­sure, towards the west. The vapour cloud thick­ens from 1m deep to about 2m deep and soon flows out in all direc­tions.

0550 to 0600 GMT: The pump­ing rate down the pipeline to Tank 912 grad­u­al­ly ris­es to around 890 cubic metres an hour, and the vapour cloud starts flow­ing off-site.

0601 GMT: The first explo­sion occurs, and fur­ther explo­sions fol­low, even­tu­al­ly engulf­ing more than 20 large stor­age tanks.

Terror in the subway

Don’t pan­ic — I nar­row­ly missed being on one of the trains attacked in the Lon­don bombs yes­ter­day. The only day I decide to ven­ture into Lon­don, just my luck…

So, it’s 9:00am and I’m on the under­ground train from Epping to Liv­er­pool St sta­tion, where I was due to catch anoth­er under­ground on the cir­cle line. Then I hear it’s not stop­ping until two stops lat­er. “Great”, I think — anoth­er british rail dis­as­ter. I get off, mild­ly con­cerned at the num­ber of peo­ple, and make my way along The Strand, towards Buck­ing­ham Palace. I notice a bunch of peo­ple inside an elec­tron­ics store (Dixons), star­ing at a TV. Pret­ty weird, so I walk in to see if it’s relat­ed to my missed stop. That’s when I realise there’s at least two fatal­i­ties, and pos­si­bly a bomb. I con­tin­ue walk­ing, and the ICA (Insti­tute of Con­tem­po­rary Arts) is closed due to a secu­ri­ty scare, so across the road I go to St James’ Park for the Liv­ing Muse­um (yeah, not a big war fan, but the Bletch­ley Park Trust — i.e. cryp­tog­ra­phers dur­ing the war — had a stand). Very bizarre, police with Uzi’s out­side, search­ing every­one. Start to get a bit weird­ed out, but hey — per­haps that’s life in Lon­don.

All the war pro­pa­gan­da was get­ting to me, so I left to see if the Tate muse­um of mod­ern art was open — on the way, stop­ping at the Lon­don Eye (closed), Dali muse­um (dit­to), and by this time get­ting a lit­tle wor­ried. So, into the pub at the Nam­co muse­um (it had a vide­owall), where I felt increas­ing­ly sur­re­al. Looked around to see who I would be strand­ed with if all trans­port was out for the day. Tried for the umpteenth time to get a text mes­sage to my uncle, who was (I thought) the only one who knew I was in Lon­don. Tony is a bit of a wor­ri­er, and had vir­tu­al­ly organ­ised my trip to the last detail — includ­ing the trip on the cir­cle line — and so would be out of his tree with wor­ry. Final­ly man­aged to get a mes­sage through, by this time my broth­er Lee and moth­er (as well as a friend in Ire­land) had seen the news and won­dered if I was ok.

I am.

Spent the next few hours wan­der­ing aim­less­ly around cen­tral Lon­don — it was hon­est­ly like A Qui­et Earth (with Bruno Lawrence) or 28 Days Lat­er — no cars at all, no bus­es — until the ear­ly after­noon not even many peo­ple. When I realised it might take quite a while to get back, I made my way to Fenchurch St sta­tion, where about 2000 peo­ple gath­ered in front of the build­ing — it was going to be hours before I even got inside.

The weird­est thing about the whole expe­ri­ence was the jun­gle grapevine — every­one had cell­phones stuck to their ears, and all you had to do was lis­ten to snip­pets of con­ver­sa­tion and you would be instant­ly updat­ed, with­out even hav­ing to speak. That’s how I learned Liv­er­pool St was still open, so I blind­ly made my way there (half the time walk­ing with the throng, the oth­er half try­ing to go against the flow).

Made my way to Liv­er­pool St, and pret­ty much straight away got on the first train that looked like it was head­ing back towards Epping — which turned out to be a first class express ser­vice. All the trains were free at this stage. Texted my Uncle to pick me up from Har­low, how­ev­er I soon found out there was a secu­ri­ty alert there, so could­n’t get off until Stanst­ed Air­port (more armed police). Final­ly man­aged to meet him at the arrivals ter­mi­nal, where many strand­ed trav­ellers could­n’t get back into Lon­don because all the trains were can­celled.

All in all, a very event­ful day.