AI is in the air. It can low­er your costs, en­hance your cus­tomer ex­pe­ri­ence, speed up your re­sponse, and in­crease your rev­enue. It is no longer for large firms; like many in­for­ma­tion tech­nolo­gies to­day, it is more af­ford­able and adapt­able for any or­gan­i­sa­tion. While as an en­tre­pre­neur, you may not need in-depth knowl­edge of how al­go­rithms work, you still need to know how to use them and how they can mim­ic hu­man think­ing to give your firm the edge.

The fol­low­ing case study was gen­er­at­ed by AI and edit­ed for this ar­ti­cle to show how it can im­pact a small busi­ness.

Light Speed Lo­gis­tics, a 25-per­son de­liv­ery firm, strug­gled with ris­ing costs and slow cus­tomer re­sponse times. The own­er ap­point­ed an “AI cham­pi­on” to test tools against three bot­tle­necks: rout­ing, cus­tomer queries and re­port­ing. They in­tro­duced an AI as­sis­tant to au­to-draft emails and in­voic­es, a chat­bot to han­dle sim­ple de­liv­ery ques­tions and an AI-pow­ered an­a­lyt­ics tool to flag late-pay­ing ac­counts. With­in 90 days, cus­tomer re­sponse times fell by 40 per cent, over­time costs dropped by 15 per cent and the team re­claimed around 10 hours a week from man­u­al pa­per­work. Staff re­port­ed less burnout and more time for com­plex cus­tomer is­sues and sales.

While I can write a fic­ti­tious case, it saves me time and al­lows me to fo­cus on ar­eas that AI can­not—my life­time of ex­pe­ri­ences as an en­tre­pre­neur­ship ed­u­ca­tor, much of which is not in the pub­lic do­main, which it can­not find. How­ev­er, AI can do more than most en­tre­pre­neurs think. In fact, in the ear­ly days, I was us­ing AI with­out even know­ing it. I was us­ing Gram­marly to im­prove my writ­ing and give it a pol­ished look. Some­time af­ter, I went back to the ar­ti­cle and no­ticed that the soft­ware was im­prov­ing it again. It did not dawn on me that it was learn­ing from mil­lions of users.

Fo­cus on busi­ness out­comes, not tech­nol­o­gy

Do not start any­where. Think about your goals and build out your strat­e­gy. You might be a start­up and want tech­nol­o­gy to help test the va­lid­i­ty of your idea. You may have guessed that wealthy, time-poor pro­fes­sion­als want help in rais­ing their kids. AI can re­search how to de­vel­op this idea in­to a busi­ness and iden­ti­fy some of the com­pe­ten­cies you need.

If you are in­ter­est­ed in scal­ing your busi­ness, how do the al­go­rithms show the pos­si­bil­i­ties? In our AI-gen­er­at­ed case, au­to­mat­i­cal­ly gen­er­at­ed emails can ex­plore rev­enue po­ten­tial. You might want to re­duce ex­pens­es, and the Light Speed case demon­strates how to low­er over­time costs and re­duce pa­per­work.

Of­ten, you have buy­ing cus­tomers, but your share of their pur­chas­es is less than it should be. The soft­ware can iden­ti­fy sim­i­lar client types and sug­gest cross-sell­ing or up­selling op­por­tu­ni­ties your team may have over­looked. Even your cus­tomers may not recog­nise that you have the so­lu­tion, and you may not ei­ther. AI can bridge that gap. For ex­am­ple, I will use my knowl­edge of fam­i­ly busi­ness man­age­ment to con­sult with wealthy fam­i­lies on suc­ces­sion and gov­er­nance. Still, on the flip side, as an in­sid­er, I can de­vel­op a mar­ket­ing strat­e­gy for fi­nan­cial prod­uct sell­ers. AI can re­search the cur­rent strate­gies of wealth man­age­ment firms, craft some sug­ges­tions, and launch a new prod­uct.

Map your cur­rent process­es

Look at what you do from be­gin­ning to end. In the mar­ket­ing func­tion, you may be spend­ing too much time on con­tent cre­ation. Let’s say you have email cam­paigns; you can use AI to gen­er­ate con­tent to ex­cite your clients about a new prod­uct or to short­en the cre­ation process for your LinkedIn newslet­ter. Re­mem­ber that you need two hats in con­tent cre­ation: be an au­thor and an ed­i­tor. They are both im­por­tant and rec­og­nize AI strengths and weak­ness­es, as well as yours.

Some clients re­quire a long pro­pos­al, and your AI, if prop­er­ly trained, can do a good draft. Some con­sul­tants spend a con­sid­er­able amount of time on this ear­ly stage of the process, with a low chance of suc­cess. Tur­ing out more well-craft­ed pro­pos­als in a short­er cy­cle can in­crease your firm’s top line.

In HR, AI can turn out learn­ing ma­te­ri­als for on­board­ing and job de­scrip­tions. It can as­sist with in­dus­try-spe­cif­ic com­pen­sa­tion plans. It can al­so im­prove your web­site by in­cor­po­rat­ing his­tor­i­cal and bi­o­graph­i­cal in­for­ma­tion, as well as a chat­bot.

Iden­ti­fy your first AI use

Af­ter you have planned out your AI strat­e­gy, the stage is ex­e­cu­tion. This is where many things can go off track. Avoid try­ing to “do AI every­where” at once, as it is easy to get over­whelmed, as some shop­pers do with ex­treme choic­es. Start with a 1-3 high-im­pact use case.

First, do you have enough da­ta or knowl­edge to sup­port an ac­tiv­i­ty to make it fea­si­ble? My AI is em­bed­ded in my soft­ware pack­age, so it can an­a­lyze my web­site and sug­gest im­prove­ments. With the right prompts, I can get sug­ges­tions for my two newslet­ters.

Sec­ond, try not to over­haul your sys­tems. But start small and learn what the tech can do. You can de­vel­op in­for­ma­tion about a client’s prob­lem that will con­vince them you have the ex­per­tise. Your web­site can rec­om­mend some prod­ucts or pro­vide more in­for­ma­tion. In­sert a call to ac­tion (CTA) with a link to a busi­ness de­vel­op­ment per­son, and in­clude a thank-you note with more read­ing. Your back end should in­di­cate whether fol­low-up oc­curred.

Third, tie the change to cost sav­ings, rev­enue or risk re­duc­tion. In the Light Speed Lo­gis­tics case, it can re­duce its re­sponse time, col­lect rev­enues faster, and im­prove cash flow. By get­ting on top of their re­ceiv­ables, they can re­duce doubt­ful ac­counts.

De­vel­op­ing an AI strat­e­gy is not about build­ing com­plex sys­tems but about align­ing it with clear, mea­sur­able goals that en­hance per­for­mance. Iden­ti­fy a few ar­eas in which your busi­ness can im­prove and de­sign mea­sur­able ex­per­i­ments to test their va­lid­i­ty. Fi­nal­ly, when these ex­per­i­ments work, scale them and mon­i­tor their per­for­mance against your goals with hu­man in­sight. Re­mem­ber, it’s your busi­ness, and AI is an in­tel­li­gent ma­chine, so al­ways be in con­trol; you’re the boss, not AI.

Saj­jad Hamid is an SME & Fam­i­ly Busi­ness Ad­vis­er, Fel­low of the Fam­i­ly Firm In­sti­tute. He can be con­tact­ed at en­tre­pre­neurt­nt@gmail.com