print ads , 200 radio scripts , 200 marketing activation plans , and 200 digital campaign case studies . I would even weave in some that are in different ethnic languages but performed at a national level . The goal here would be to sample the best of what Kenya ' s marketing , design , copy , and creative minds have developed over the years .
As Nendo is both a research and marketing agency in one , marketing effectiveness would have to be featured . It would involve scrutinizing the campaigns in which the creative work delivered measurable business results : salience , penetration , fame , brand building , sales activation , excess Share of Voice , and the entire bouquet of marketing effectiveness terms and concepts . It might begin to separate the top 10 % of campaigns that were impressively creative and , importantly , commercially successful in shifting perceptions and behavior . I ' ve always wished Nendo would build a database akin to the UK ' s Institute of Practitioners of Advertising ( IPA ) and their Data Bank of marketing effectiveness case studies . If Kenya were to one day set up its own Effies , an award ceremony built to focus on how marketing drives business results , it would bring a tear to my eye .
The next step would be explaining what worked about these campaigns in as many words as necessary . Words to explain the sheng , slang , Swahili , vernacular , and cultural contexts . Creating as many words as possible to describe the creative output and rationales . Hypothesizing why they delivered and made sense . Keep in mind that Large Language Models ( LLMs ) work with words . When done with this , decisions might have context , such as why certain voices were chosen or visual creative choices were taken . Anything to feed the AI engine with Kenyan context . Still images , photo compositions , dates , times , locations , and more would serve as fodder for this .
Then , we ' d need to chronicle why the work landed and struck a chord with Kenyans . More words . Feeding the machine for the reason that it is " Mimi ni Member " and not " Mimi ni Memba ." To find a way to distil strange idiosyncrasies of the creative geniuses that midwifed and birthed gamechanging marketing campaign ideas .
We can ' t replicate the work of the idea coming to a copywriter in the middle of a busy commute in the back of a matatu , music flowing through earphones , watching the blur of oncoming traffic in Nairobi ' s smoggy haze , or replicating
Even if the AI were to work , I would need help with the idea that a chatbot could do what creative professionals in Nairobi do . Having seen my team use ChatGPT +, Midjourney , Perplexity , and other tools over the past few months , AI is exciting , but I have yet to find the ‘ soul ’ behind it .
the copywriter on a smoke break fiddling with a bug on the floor on a rooftop in Nairobi ' s uncannily cold and breezy Julys .
Or the awkward silence that befalls a boardroom as the main idea that everyone worked on for the past two weeks falls flat , and the creative director has to adlib by stepping up to the whiteboard and quenching their dry throat with a gulp of saliva before waving a whiteboard marker and cracking the brief in a moment of unscripted genius .
It would be impossible to capture lighting in a bottle , particularly in a city with as many contrasts , coincidences , and curiosities as Nairobi . I would struggle to translate a way to capture a millionth of the incalculable essence and translate this into words for the ' machine .'
At some point , we ' d run out of words to add and just need to ‘ turn it on ’. Akin to Frankenstein , we ' d need to test how this Kenyan AI Copywriter and Creative Director works . We would need a brief . And as a researcher , you can expect me to have a control group and a treatment group . The AI on its own would serve alongside a creative team ( copywriter and art director ) as the treatment group , and a regular warm-blooded art & copy team of Nairobi creative professionals would work as usual . Man vs . machine . A duel for the ages .
A final word on Mutura and Machines
Even if the AI were to work , I would need help with the idea that a chatbot could do what creative professionals in Nairobi do . Having seen my team use ChatGPT +, Midjourney , Perplexity , and other tools over the past few months , AI is exciting , but I have yet to find the ‘ soul ’ behind it .
I ' d liken it to an online conversation regarding mutura , a street food African blood sausage often made at nyama choma
( barbeque ) joints across the country . Thanks to TikTok ' s wonders , there have been videos of Asian chefs preparing a replica of the same dish , especially regarding the ingredients .
The surprising remarks from Kenyans were that it just isn ' t the same . " There ' s too much light there ," one would say - referring to studio lighting in the TikTok video compared to the dingy and dim lighting often where mutura is prepared and served locally . " The cleaning surfaces ( in the video ) are too clean ," another would add . Expressing a disappointment that the kitchen surfaces , chopping board , knives , and utensils are too sanitary , not allowing for an uncoordinated yet delicious mingling of flavors . Others would remark that if the mutura isn ' t prepared on the street with the smog , dust , and less-thansanitary conditions , there ' s a flavor and a heart-racing risk level of H-pylori that beats the purpose of the mutura itself .
Could this be why our AI-generated Kenyan creative would fail ? Kenyans are non-linear . Like Sunny Bindra said , they are peculiar . So why would the creative ideas to change their hearts , capture their minds , or move their wallets come out of a test tube laboratory - even when it is the world ' s best AI copywriter ?
Mark Kaigwa is the founder of Nendo , an award-winning digital research and marketing firm based in Nairobi that works across Africa . Nendo delivers grounded African insights and trends . Visit nendo . co . ke to learn more and download free trend reports and resources . You can commune with mark via mail at : Mark @ nendo . co . ke .