by Oguz A. Acar and Andrés Gvirtz for Harvard Business Review
Dollar Shave Club’s audacious market entry is nothing short of iconic in the marketing world. The bold startup dared to challenge the formidable razor titan, Gillette, with a video that was filmed in just one day at a mere cost of $4,500. Their strategy, marked by a mix of cheeky irreverence and simple value proposition, quickly went viral and captured interest worldwide.
However, such stories remain extremely rare. And it’s often because the behemoths have an inherent advantage. Their extensive resources — be it in technology, market intelligence, or content creation — often tip the scales in their favor. Consider this: large companies spent on average $7 million for a 30-second spot at the Super Bowl. This single spot alone is significantly greater than annual revenue of most small companies.
The results of a Deloitte survey further illustrate the scale of the spending differences. Even though smaller companies — those with revenue under $10 million — spend more on marketing in relative terms (15.2% of revenue), large companies — with revenue over $10 billion — still spend an important portion (7.1%), leading to a stark imbalance in absolute marketing expenditures.
Generative AI Tools for SMEs
But the times are changing. Generative AI is arming smaller companies with once unattainable capabilities, which, if used strategically, may make the playing field more level. Take a key aspect of marketing for example: content creation. Producing high-quality, consistent, and engaging content has often been a luxury beyond the reach of small and midsized enterprises (SMEs). But now, even a neighborhood boutique can leverage generative AI to create vivid product descriptions and imagery, as well as ads, all while keeping costs at bay.
Consider platforms like Jasper that are focusing on marketing content creation. These tools help create engaging posts, tailor-made for various social media platforms, at a fraction of the cost and time. Likewise, platforms like Canva and Adobe Firefly focus on images, allowing SMEs to produce visually appealing content without the need for expensive graphic design resources.
While these platforms often advertise the multinational enterprise clients that use their services, dig deeper into some of the 100,000 companies adopting these technologies, and you can see the real beneficiaries. For example, reviews for Jasper include a musician who launched his first single, an immigrant and software engineer who no longer feels language is a barrier, and small companies who suddenly feel like they can have professional search engine optimization (SEO). Similarly, when analyzing the reviews of Neuroflash, a European AI content suite, we noted that 354 out of 398 five-star reviews were from smaller companies; it seems that those who are most enthusiastic about these innovations are SMEs who have traditionally lacked consistent access to services like copyrighting and graphic design.
Text and Image Generation
The rise of open source models for text and image generation is also significant trend, as this makes large language models (LLMs) increasingly cost-effective and flexible. For example, Mistral AI has released a free open-source LLM via a torrent link, which can be downloaded and run on an ordinary PC without the need for expensive computing resources. These open-source models can also be modified to align with business objectives and needs. For example, in a recent study, researchers developed an ad generation algorithm based on open-source generative AI tools. Impressively, these ads not only maintained intended brand personality but also outperformed actual industry ads in common performance metrics.
Recent advances in generative AI go well beyond text and image. Google’s recent MusicLM can generate music from text, for example, in response to a prompt like: “the main soundtrack of an arcade game.” Likewise, OpenAI released an open-source algorithm capable of composing music, including vocal tracks. Indeed, AI-generated music has already found its ways into advertisements of perfumes and cars. The highlighted examples are about prominent players but the value proposition is even stronger for SMEs given the steep costs associated with music royalties. These new tools offer them a new avenue for affordable, high-quality musical content.
AI-Generated Video
Further change is on the horizon, with AI-generated video content creation rapidly evolving. Although currently trailing behind other media forms in development, the progress and investment in this sector area promising. For example, Synthesia has attracted significant backing by industry giants like Nvidia. Similarly, initiatives like Runway’s short film contest illustrate the future potential for AI in generating high-quality and artistic video content. This momentum suggests that generating marketing videos might soon be within reach for SMEs.
These videos can even be created truly globally, with platforms like Heygen breaking down language barriers by accurately and realistically translating video content. For instance, an Italian artisanal cheese maker can now easily target a global audience by translating promotional material into various languages, like German and Hindi. This capability, a daunting challenge even for large corporations due to resource constraints, is a step towards democratizing global access.
The scale advantage was not just in content creation; it was deeply entrenched in intelligence. Large corporations had the means and resources to conduct thorough research, gaining granular insights into markets and consumer behavior. Smaller companies, in contrast, often lacked the luxury of advanced consumer surveys and specialized analytics.
Improved Customer Insights
Today, AI can already sift through vast data sources and generate market and customer insights. Whether it’s analyzing online sentiment, assessing competitor positioning, or understanding visual preferences, AI can amplify analytical capabilities of SMEs. For example, an SME trying to understand customer purchasing patterns can use tools like OpenAI’s Code Interpreter to conduct advanced analyses. By entering simple prompts like “Show me monthly sales trends by product” or “identify the key drivers of consumer sentiments,” SMEs can navigate through different methodological options and get a custom code for data visualization and analysis, even if they lack proficiency in data analytics languages like Python or R.
Generative AI is not only instrumental in analyzing data but also in its generation. Research, including our own on over 20,000 chatbots, highlights that LLMs can be used to simulate human participants. However, it is important to be cautious, as the findings might not be broadly applicable across different target audiences. Researchers have, for example, showed that LLM outputs are more aligned with responses from WEIRD (Western, Educated, Industrialized, Rich, and Democratic) societies. Yet, as with most other findings on LLMs, this seems to depend on the chosen model and prompts. For instance, one study found that GPT-4 could accurately reflect personality traits differences between U.S. and South Korean individuals, when prompted in the respective languages and instructed to mimic these audiences.
Custom ChatGPT Chatbots
Other significant changes are also on the horizon. Soon, we might see AI agents that can undertake specific business tasks. Consider the recent release of OpenAI’s GPT models which allow users to develop custom chatbots tailored to specific functions, such as brand management, technology advising, financial consulting, or event scheduling. Within just 10 days since its release, we have already seen these custom GPTs taking on roles like strategy consultants, travel planners, and marketing assistants. While there is room for growth and improvement in their capabilities, they offer a preview of a future where diverse AI agents fulfill distinct roles and interact in different ways.
The real potential lies in integrating all these advances in a multi-modal and multi-agent AI system. Imagine a marketing team using an AI system that harmoniously blends different functions such as content generation, branding, and data analytics. For example, our Italian artisanal cheese maker can use these advances to elevate every aspect of its online presence, from content and imagery to language and design.
All in all, multi-modal and multi-agent generative AI has the potential to close the content, insight, and technology gaps that large corporations typically have over their smaller counterparts. This shift presents a unique opportunity for SMEs, whose inherent agility gives them an edge in adopting and innovating with AI. It’s not far-fetched, then, to foresee a future marketplace where the depth of corporate resources is not the unequivocal determinant of success.
Leave a Reply